bibliography.bib

@misc{github-repo,
  title = {GitHub Repository for bachelor thesis},
  author = {Lina Wilske},
  url = {https://github.com/linaScience/ba-implementation}
}
@article{MLBasedHandoverPrediction2022,
  title = {{{ML-Based Handover Prediction}} and {{AP Selection}} in {{Cognitive Wi-Fi Networks}}},
  author = {Khan, Muhammad Asif and Hamila, Ridha and Gastli, Adel and Kiranyaz, Serkan and {Al-Emadi}, Nasser Ahmed},
  year = {2022},
  month = aug,
  journal = {Journal of Network and Systems Management},
  volume = {30},
  number = {4},
  pages = {72},
  issn = {1573-7705},
  doi = {10.1007/s10922-022-09684-2},
  urldate = {2023-04-26},
  abstract = {Device mobility in dense Wi-Fi networks offers several challenges. Two well-known problems related to device mobility are handover prediction and access point selection. Due to the complex nature of the radio environment, analytical models may not characterize the wireless channel, which makes the solution of these problems very difficult. Recently, cognitive network architectures using sophisticated learning techniques are increasingly being applied to such problems. In this paper, we propose data-driven machine learning (ML) schemes to efficiently solve these problems in wireless LAN (WLAN) networks. The proposed schemes are evaluated and results are compared with traditional approaches to the aforementioned problems. The results report significant improvement in network performance by applying the proposed schemes. The proposed scheme for handover prediction outperforms traditional methods i.e. received signal strength method and traveling distance method by reducing the number of unnecessary handovers by 60\% and 50\% respectively. Similarly, in AP selection, the proposed scheme outperforms the strongest signal first and least loaded first algorithms by achieving higher throughput gains up to 9.2\% and 8\% respectively.},
  langid = {english},
  keywords = {Access point selection,Cognitive networks,Handover,Machine learning,Throughput,Wi-Fi},
  file = {/Users/lina/Zotero/storage/EP66AE87/Khan et al. - 2022 - ML-Based Handover Prediction and AP Selection in C.pdf}
}

@article{bohannonComfortableMaximumWalking1997,
  title = {Comfortable and Maximum Walking Speed of Adults Aged 20-79 Years: Reference Values and Determinants},
  shorttitle = {Comfortable and Maximum Walking Speed of Adults Aged 20-79 Years},
  author = {Bohannon, R. W.},
  year = {1997},
  month = jan,
  journal = {Age and Ageing},
  volume = {26},
  number = {1},
  pages = {15--19},
  issn = {0002-0729},
  doi = {10.1093/ageing/26.1.15},
  abstract = {OBJECTIVES: to establish reference values for both comfortable and maximum gait speed and to describe the reliability of the gait speed measures and the correlation of selected variables with them. DESIGN: descriptive and cross-sectional. METHODS: subjects were 230 healthy volunteers. Gait was timed over a 7.62 m expanse of floor. Actual and height normalized speed were determined. Lower extremity muscle strength was measured with a hand-held dynamometer. RESULTS: mean comfortable gait speed ranged from 127.2 cm/s for women in their seventies to 146.2 cm/s for men in their forties. Mean maximum gait speed ranged from 174.9 cm/s for women in their seventies to 253.3 cm/s for men in their twenties. Both gait speed measures were reliable (coefficients {$>$} or = 0.903) and correlated significantly with age (r {$>$} or = -0.210), height (r {$>$} or = 0.220) and the strengths of four measured lower extremity muscle actions (r = 0.190-0.500). The muscle action strengths most strongly correlated with gait speed were nondominant hip abduction (comfortable speed) and knee extension (maximum speed). CONCLUSIONS: these normative values should give clinicians a reference against which patient performance can be compared in a variety of settings. Gait speed can be expected to be reduced in individuals of greater age and of lesser height and lower extremity muscle strength.},
  langid = {english},
  pmid = {9143432},
  keywords = {Adult,Aged,Aging,Female,Gait,Geriatric Assessment,Humans,Male,Middle Aged,Muscle Contraction,Physical Fitness,Reference Values,Walking},
  file = {/Users/lina/Zotero/storage/9TXE3JX9/Bohannon - 1997 - Comfortable and maximum walking speed of adults ag.pdf}
}

@misc{binary-classification,
	author = {Karabiber, Faith},
	title = {Binary Classification},
	url = {https://www.learndatasci.com/glossary/binary-classification/},
	urldate={2023-09-03},
}

@misc{tanh-lstm-default,
	title = {LSTM layer},
	author = {keras},
	url = {https://keras.io/api/layers/recurrent_layers/lstm/},
	urldate ={2023-09-03},
}

@misc{yuHyperParameterOptimizationReview2020,
  title = {Hyper-{{Parameter Optimization}}: {{A Review}} of {{Algorithms}} and {{Applications}}},
  shorttitle = {Hyper-{{Parameter Optimization}}},
  author = {Yu, Tong and Zhu, Hong},
  year = {2020},
  month = mar,
  number = {arXiv:2003.05689},
  eprint = {2003.05689},
  primaryclass = {cs, stat},
  publisher = {{arXiv}},
  doi = {10.48550/arXiv.2003.05689},
  urldate = {2023-09-03},
  keywords = {Computer Science - Machine Learning,Statistics - Machine Learning},
  file = {/Users/lina/Zotero/storage/R47SX7DZ/Yu and Zhu - 2020 - Hyper-Parameter Optimization A Review of Algorith.pdf;/Users/lina/Zotero/storage/UUBEXFFX/2003.html}
}


@article{mirzaMachineLearningApproach2007,
  title = {A Machine Learning Approach to {{TCP}} Throughput Prediction},
  author = {Mirza, Mariyam and Sommers, Joel and Barford, Paul and Zhu, Xiaojin},
  year = {2007},
  month = jun,
  journal = {ACM SIGMETRICS Performance Evaluation Review},
  volume = {35},
  number = {1},
  pages = {97--108},
  issn = {0163-5999},
  doi = {10.1145/1269899.1254894},
  urldate = {2023-04-27},
  file = {/Users/lina/Zotero/storage/BF64NSXV/Mirza et al. - 2007 - A machine learning approach to TCP throughput pred.pdf}
}


@inproceedings{handover-assisted-by-gps,
  title = {{{IEEE}} 802.11 {{Handovers Assisted}} by {{GPS Information}}},
  booktitle = {2006 {{IEEE International Conference}} on {{Wireless}} and {{Mobile Computing}}, {{Networking}} and {{Communications}}},
  author = {Montavont, J. and Noel, T.},
  year = {2006},
  month = jun,
  pages = {166--172},
  issn = {2160-4894},
  doi = {10.1109/WIMOB.2006.1696358},
  abstract = {IEEE 802.11 networks are now very common and are present in various locations. While roaming through access points, a mobile node is often required to perform a link layer handover. This mechanism causes user-interceptable connection loss and breaks in time-sensitive communication, especially if a network layer handover follows the link layer handover. Many solutions attempting to improve this process have been proposed but only a few use geolocation systems in the management of the handover. In this article, we present a new method to enhance both link layer and network layer handovers using geolocation information provided by a GPS system. The idea behind our algorithm is to predict the next mobile node point of attachment and the associated sub-network using the position of the mobile nodes. This method has been implemented using the new Mobile IP daemon for GNU/Linux operating system and evaluated through two scenarios},
  keywords = {Broadcasting,Global Positioning System,Linux,Mobile communication,Operating systems,Prediction algorithms,Probes,Roaming,Telecommunication traffic,Throughput},
  file = {/Users/lina/Zotero/storage/7A2U2Y55/1696358.html}
}


@misc{kaggle,
  title = {Kaggle: {{Your Home}} for {{Data Science}}},
  urldate = {2023-07-23},
  url = {https://www.kaggle.com/},
  file = {/Users/lina/Zotero/storage/2AYKSBEW/www.kaggle.com.html}
}

@misc{handoff_performance_issues,
	url = {https://mentor.ieee.org/802.11/dcn/22/11-22-1874-02-0wng-roaming-handoff-time-reduction-to-improve-user-experience.pptx},
	url_date = {2023-09-01},
	author = {Hu, Xiaokun and Zheng, Lei and Chen, Jianxiang and Sun, Lingwu and Zhu, Lihua}
}

@misc{rssi_calculation,
	url = {https://developer.android.com/reference/android/net/wifi/WifiManager#calculateSignalLevel(int)},
	url_date = {2023-09-01},
	author = {Android Developers}
}

@misc{IndoorLocationNavigation,
  title = {Indoor {{Location}} \& {{Navigation}} | {{Kaggle}}},
  urldate = {2023-07-11},
  url = {https://www.kaggle.com/competitions/indoor-location-navigation},
  file = {/Users/lina/Zotero/storage/FRT9ZVVT/indoor-location-navigation.html}
}

@article{EffectsSlidingWindow2022,
  title = {Effects of Sliding Window Variation in the Performance of Acceleration-Based Human Activity Recognition Using Deep Learning Models},
  author = {{Ja{\'e}n-Vargas}, Milagros and Leiva, Karla Miriam Reyes and Fernandes, Francisco and Gon{\c c}alves, S{\'e}rgio Barroso and Silva, Miguel Tavares and Lopes, Daniel Sim{\~o}es and Olmedo, Jos{\'e} Javier Serrano},
  year = {2022},
  month = aug,
  journal = {PeerJ Computer Science},
  volume = {8},
  pages = {e1052},
  publisher = {{PeerJ Inc.}},
  issn = {2376-5992},
  doi = {10.7717/peerj-cs.1052},
  urldate = {2023-08-27},
  abstract = {Deep learning (DL) models are very useful for human activity recognition (HAR); these methods present better accuracy for HAR when compared to traditional, among other advantages. DL learns from unlabeled data and extracts features from raw data, as for the case of time-series acceleration. Sliding windows is a feature extraction technique. When used for preprocessing time-series data, it provides an improvement in accuracy, latency, and cost of processing. The time and cost of preprocessing can be beneficial especially if the window size is small, but how small can this window be to keep good accuracy? The objective of this research was to analyze the performance of four DL models: a simple deep neural network (DNN); a convolutional neural network (CNN); a long short-term memory network (LSTM); and a hybrid model (CNN-LSTM), when variating the sliding window size using fixed overlapped windows to identify an optimal window size for HAR. We compare the effects in two acceleration sources': wearable inertial measurement unit sensors (IMU) and motion caption systems (MOCAP). Moreover, short sliding windows of sizes 5, 10, 15, 20, and 25 frames to long ones of sizes 50, 75, 100, and 200 frames were compared. The models were fed using raw acceleration data acquired in experimental conditions for three activities: walking, sit-to-stand, and squatting. Results show that the most optimal window is from 20\textendash 25 frames (0.20\textendash 0.25s) for both sources, providing an accuracy of 99,07\% and F1-score of 87,08\% in the (CNN-LSTM) using the wearable sensors data, and accuracy of 98,8\% and F1-score of 82,80\% using MOCAP data; similar accurate results were obtained with the LSTM model. There is almost no difference in accuracy in larger frames (100, 200). However, smaller windows present a decrease in the F1-score. In regard to inference time, data with a sliding window of 20 frames can be preprocessed around 4x (LSTM) and 2x (CNN-LSTM) times faster than data using 100 frames.},
  langid = {english},
  file = {/Users/lina/Zotero/storage/44BIVPKE/Jaén-Vargas et al. - 2022 - Effects of sliding window variation in the perform.pdf}
}

@misc{IndoorNavigationUnderstanding,
  title = {Indoor Navigation: Complete Data Understanding},
  shorttitle = {{{Indoor Navigation}}},
  urldate = {2023-04-25},
  url = {https://kaggle.com/code/andradaolteanu/indoor-navigation-complete-data-understanding},
  langid = {english},
}

@misc{GitHubComp,
	title = {indoor-location-navigation-20},
	url = {https://github.com/location-competition/indoor-location-competition-20},
	urldate = {2023-07-31},
}
@misc{mnist,
	title = {The MNIST Database},
	url = {http://yann.lecun.com/exdb/mnist/},
	urldate = {2023-08-22}
}
@article{TSC,
	author = {Ismail Fawaz, Hassan and Forestier, Germain and Weber, Jonathan and Idoumghar, Lhassane and Muller, Pierre-Alain},
	doi = {10.1007/s10618-019-00619-1},
	issn = {1384-5810, 1573-756X},
	journal = {Data Mining and Knowledge Discovery},
	language = {en},
	month = jul,
	number = {4},
	pages = {917--963},
	shorttitle = {Deep learning for time series classification},
	title = {Deep learning for time series classification: a review},
	url = {https://link.springer.com/10.1007/s10618-019-00619-1},
	urldate = {2023-08-06},
	volume = {33},
	year = {2019},
	bdsk-url-1 = {https://link.springer.com/10.1007/s10618-019-00619-1},
	bdsk-url-2 = {https://doi.org/10.1007/s10618-019-00619-1}
}
@article{multivariate-lstm-indoor-outdoor,
  title = {Deep-{{Learning-Based Multivariate Time-Series Classification}} for {{Indoor}}/{{Outdoor Detection}}},
  author = {Bakirtzis, Stefanos and Qiu, Kehai and Wassell, Ian and Fiore, Marco and Zhang, Jie},
  year = {2022},
  month = dec,
  journal = {IEEE Internet of Things Journal},
  volume = {9},
  pages = {24529--24540},
  issn = {2327-4662},
  doi = {10.1109/JIOT.2022.3190555},
  abstract = {Recently, the topic of indoor outdoor detection (IOD) has seen its popularity increase, as IOD models can be leveraged to augment the performance of numerous Internet of Things and other applications. IOD aims at distinguishing in an efficient manner whether a user resides in an indoor or an outdoor environment, by inspecting the cellular phone sensor recordings. Legacy IOD models attempt to determine a user's environment by comparing the sensor measurements to some threshold values. However, as we also observe in our experiments, such models exhibit limited scalability, and their accuracy can be poor. Machine learning (ML)-based IOD models aim at removing this limitation, by utilizing a large volume of measurements to train ML algorithms to classify a user's environment. Yet, in most of the existing research, the temporal dimension of the problem is disregarded. In this article, we propose treating IOD as a multivariate time-series classification (TSC) problem, and we explore the performance of various deep learning (DL) models. We demonstrate that a multivariate TSC approach can be used to monitor a user's environment, and predict changes in its state, with greater accuracy compared to conventional approaches that ignore the feature variation over time. Additionally, we introduce a new DL model for multivariate TSC, exploiting the concept of self-attention and atrous spatial pyramid pooling. The proposed DL multivariate TSC framework exploits only low power consumption sensors to infer a user's environment, and it outperforms state-of-the-art models, yielding a higher accuracy combined with a smaller computational cost.},
  keywords = {Biological system modeling,Computational modeling,Deep learning (DL),indoor\textendash outdoor detection (IOD),Internet of Things,Predictive models,seamless navigation,self-attention,Time measurement,Time series analysis,time-series classification (TSC),Wireless fidelity},
  file = {/Users/lina/Zotero/storage/CVRTYP59/Bakirtzis et al. - 2022 - Deep-Learning-Based Multivariate Time-Series Class.pdf;/Users/lina/Zotero/storage/UILCCTWQ/9828386.html}
}
@article{bourjandiPredictingUserMovement2022,
  title = {Predicting User's Movement Path in Indoor Environments Using the Stacked Deep Learning Method and the Fuzzy Soft-Max Classifier},
  author = {Bourjandi, Masoumeh and {Yadollahzadeh-Tabari}, Meisam and GolsorkhtabariAmiri, Mehdi},
  year = {2022},
  journal = {IET Signal Processing},
  volume = {16},
  number = {5},
  pages = {546--561},
  issn = {1751-9683},
  doi = {10.1049/sil2.12125},
  urldate = {2023-08-27},
  file = {/Users/lina/Zotero/storage/HE5GKCM9/Bourjandi et al. - 2022 - Predicting user's movement path in indoor environm.pdf;/Users/lina/Zotero/storage/RWVJVQPB/sil2.html}
}

@inproceedings{hmm-movement-prediction,
  title = {Movement {{Prediction}} in {{Wireless Networks Using Mobility Traces}}},
  booktitle = {2010 7th {{IEEE Consumer Communications}} and {{Networking Conference}}},
  author = {Prasad, Pratap S. and Agrawal, Prathima},
  year = {2010},
  month = jan,
  pages = {1--5},
  issn = {2331-9860},
  doi = {10.1109/CCNC.2010.5421613},
  abstract = {Wireless user-mobility prediction has been investigated from various angles to improve network performance. Student populations in campuses, pedestrian and vehicular movement in urban areas, etc have been studied by cell phone and mobility management researchers to address issues in quality of service (QoS), seamless session handoffs, etc. Access to information such as user movement times, direction, speed, etc provides an opportunity for networks to efficiently manage resources to satisfy user needs. Towards this goal, we propose a generic framework to approach the problem of mobility prediction using hidden Markov models (HMM). This method can be used to model hidden parameters in the models. We propose a way to extract user movement information from a real dataset, train a HMM using this data and make predictions using the HMM. This model can successfully predict long sequences of a mobile user's path from observed sequences and also uses successive sequences of observed data to train its learning parameters to enhance prediction accuracy. Furthermore, we show that this model is very generic and can be suited to make predictions using the same information from the perspective of the access point or the mobile node.},
  keywords = {Accuracy,Cellular phones,Data mining,Hidden Markov models,Mobile radio mobility management,Predictive models,Quality of service,Resource management,Urban areas,Wireless networks},
  file = {/Users/lina/Zotero/storage/V2YYP5ZG/Prasad and Agrawal - 2010 - Movement Prediction in Wireless Networks Using Mob.pdf;/Users/lina/Zotero/storage/6ZZ4HCCK/5421613.html}
}
@article{lstm-hochreiter,
	title = {Long Short-Term Memory},
	author = {Hochreiter, Sepp and Schmidhuber, Jürgen},
	year = {1997},
	url = {https://papers.baulab.info/Hochreiter-1997.pdf},
}
@inproceedings{comparison-lstm-mlp,
	abstract = {Neural networks is considered one of the most developed concept in artificial intelligence, due to its ability to solve complex computational tasks, and its efficiency to find solutions. There is a wide range of applications that adopt this technique, one of which is in the financial investment issues. This paper presents an approach to predict stock market ratios using artificial neural networks. It considers two different techniques- BPA-MLP and LSTM-RNN- their potential, and their limitations. Tests were conducted on different data sets, such as FacebookTM stocks, GoogleTM stocks, and BitcoinTM stocks. We achieve a best case accuracy of 97\% for MLP algorithm, and 99.5\% for LSTM algorithm. While the results appear to be promising, a web interface is presented in order to accept a certain amount of money, and accordingly checks the best stock to invest in.},
	author = {Achkar, Roger and Elias-Sleiman, Fady and Ezzidine, Hasan and Haidar, Nourhane},
	booktitle = {2018 6th {International} {Symposium} on {Computational} and {Business} {Intelligence} ({ISCBI})},
	doi = {10.1109/ISCBI.2018.00019},
	keywords = {Biological neural networks, Neurons, Recurrent neural networks, Computer architecture, Training, Logic gates, Stock markets, Back propagation Algorithm, Multi-layer Perceptron, Long Short-Term Memory, Recurrent Neural Networks},
	month = aug,
	pages = {48--51},
	title = {Comparison of {BPA}-{MLP} and {LSTM}-{RNN} for {Stocks} {Prediction}},
	year = {2018},
	bdsk-url-1 = {https://doi.org/10.1109/ISCBI.2018.00019}
}
@article{mlp-backpropagation-rumelhart,
  title={Learning representations by back-propagating errors},
  author={David E. Rumelhart and Geoffrey E. Hinton and Ronald J. Williams},
  journal={Nature},
  year={1986},
  volume={323},
  pages={533-536},
  url={https://api.semanticscholar.org/CorpusID:205001834}
}
@article{hopfield-rnn,
  title = {Neural Networks and Physical Systems with Emergent Collective Computational Abilities.},
  author = {Hopfield, J J},
  year = {1982},
  journal = {Proceedings of the National Academy of Sciences},
  volume = {79},
  number = {8},
  pages = {2554--2558},
  publisher = {{Proceedings of the National Academy of Sciences}},
  doi = {10.1073/pnas.79.8.2554},
  urldate = {2023-09-02},
  file = {/Users/lina/Zotero/storage/JYH2742V/Hopfield - 1982 - Neural networks and physical systems with emergent.pdf}
}

@book{goodfellow_deep_2016,
	abstract = {Deep learning is a form of machine learning that enables computers to learn from experience and understand the world in terms of a hierarchy of concepts. Because the computer gathers knowledge from experience, there is no need for a human computer operator to formally specify all the knowledge that the computer needs. The hierarchy of concepts allows the computer to learn complicated concepts by building them out of simpler ones; a graph of these hierarchies would be many layers deep. This book introduces a broad range of topics in deep learning. The text offers mathematical and conceptual background, covering relevant concepts in linear algebra, probability theory and information theory, numerical computation, and machine learning. It describes deep learning techniques used by practitioners in industry, including deep feedforward networks, regularization, optimization algorithms, convolutional networks, sequence modeling, and practical methodology; and it surveys such applications as natural language processing, speech recognition, computer vision, online recommendation systems, bioinformatics, and video games. Finally, the book offers research perspectives, covering such theoretical topics as linear factor models, autoencoders, representation learning, structured probabilistic models, Monte Carlo methods, the partition function, approximate inference, and deep generative models. Deep Learning can be used by undergraduate or graduate students planning careers in either industry or research, and by software engineers who want to begin using deep learning in their products or platforms. A website offers supplementary material for both readers and instructors"--Page 4 of cover},
	address = {Cambridge, Massachusetts},
	author = {Goodfellow, Ian and Bengio, Yoshua and Courville, Aaron},
	isbn = {9780262337434},
	language = {eng},
	note = {OCLC: 987005922},
	publisher = {The MIT Press},
	title = {Deep learning},
	year = {2016}
}

@book{brownleeDeepLearningTime,
  title = {Deep {{Learning}} for {{Time Series Forecasting}}},
  author = {Brownlee, Jason},
  langid = {english},
  url = {https://machinelearningmastery.com/deep-learning-for-time-series-forecasting/},
  urldate = {2023-09-03},
  file = {/Users/lina/Zotero/storage/CDZVTRI6/Brownlee - Deep Learning for Time Series Forecasting.pdf}
}

@article{ferreiraForecastingNetworkTraffic2023,
  title = {Forecasting {{Network Traffic}}: {{A Survey}} and {{Tutorial With Open-Source Comparative Evaluation}}},
  shorttitle = {Forecasting {{Network Traffic}}},
  author = {Ferreira, Gabriel O. and Ravazzi, Chiara and Dabbene, Fabrizio and Calafiore, Giuseppe C. and Fiore, Marco},
  year = {2023},
  journal = {IEEE Access},
  volume = {11},
  pages = {6018--6044},
  issn = {2169-3536},
  doi = {10.1109/ACCESS.2023.3236261},
  file = {/Users/lina/Zotero/storage/ADSGH27V/Ferreira et al. - 2023 - Forecasting Network Traffic A Survey and Tutorial.pdf;/Users/lina/Zotero/storage/HMMDIG52/stamp.html}
}
@article{meiRealtimeMobileBandwidth2022,
  title = {Realtime Mobile Bandwidth and Handoff Predictions in {{4G}}/{{5G}} Networks},
  author = {Mei, Lifan and Gou, Jinrui and Cai, Yujin and Cao, Houwei and Liu, Yong},
  year = {2022},
  month = feb,
  journal = {Computer Networks},
  volume = {204},
  pages = {108736},
  issn = {1389-1286},
  doi = {10.1016/j.comnet.2021.108736},
  urldate = {2023-09-03},
  file = {/Users/lina/Zotero/storage/CELWZIEB/Mei et al. - 2022 - Realtime mobile bandwidth and handoff predictions .pdf;/Users/lina/Zotero/storage/9T93YRMI/S1389128621005879.html}
}


@article{srivastava14a,
  author  = {Nitish Srivastava and Geoffrey Hinton and Alex Krizhevsky and Ilya Sutskever and Ruslan Salakhutdinov},
  title   = {Dropout: A Simple Way to Prevent Neural Networks from Overfitting},
  journal = {Journal of Machine Learning Research},
  year    = {2014},
  volume  = {15},
  number  = {56},
  pages   = {1929--1958},
  url     = {http://jmlr.org/papers/v15/srivastava14a.html}
}


@article{hmm-rabiner-1989,
	author = {Rabiner, L.R.},
	doi = {10.1109/5.18626},
	issn = {00189219},
	journal = {Proceedings of the IEEE},
	month = feb,
	number = {2},
	pages = {257--286},
	title = {A tutorial on hidden {Markov} models and selected applications in speech recognition},
	url = {http://ieeexplore.ieee.org/document/18626/},
	urldate = {2023-08-06},
	volume = {77},
	year = {1989},
	bdsk-url-1 = {http://ieeexplore.ieee.org/document/18626/},
	bdsk-url-2 = {https://doi.org/10.1109/5.18626}
}
@article{elman_finding_1990,
	author = {Elman, Jeffrey L.},
	doi = {10.1207/s15516709cog1402_1},
	issn = {03640213},
	journal = {Cognitive Science},
	language = {en},
	month = mar,
	number = {2},
	pages = {179--211},
	title = {Finding {Structure} in {Time}},
	urldate = {2023-08-06},
	volume = {14},
	year = {1990},
	bdsk-url-1 = {http://doi.wiley.com/10.1207/s15516709cog1402_1},
	bdsk-url-2 = {https://doi.org/10.1207/s15516709cog1402_1}
}
@inproceedings{cho_learning_2014,
	address = {Doha, Qatar},
	author = {Cho, Kyunghyun and Van Merrienboer, Bart and Gulcehre, Caglar and Bahdanau, Dzmitry and Bougares, Fethi and Schwenk, Holger and Bengio, Yoshua},
	booktitle = {Proceedings of the 2014 {Conference} on {Empirical} {Methods} in {Natural} {Language} {Processing} ({EMNLP})},
	doi = {10.3115/v1/D14-1179},
	language = {en},
	pages = {1724--1734},
	publisher = {Association for Computational Linguistics},
	title = {Learning {Phrase} {Representations} using {RNN} {Encoder}--{Decoder} for {Statistical} {Machine} {Translation}},
	url = {http://aclweb.org/anthology/D14-1179},
	urldate = {2023-08-06},
	year = {2014},
	bdsk-url-1 = {http://aclweb.org/anthology/D14-1179},
	bdsk-url-2 = {https://doi.org/10.3115/v1/D14-1179}
}
@misc{rnn_difficulties_2013,
  title = {On the Difficulty of Training {{Recurrent Neural Networks}}},
  author = {Pascanu, Razvan and Mikolov, Tomas and Bengio, Yoshua},
  year = {2013},
  month = feb,
  number = {arXiv:1211.5063},
  eprint = {1211.5063},
  primaryclass = {cs},
  publisher = {{arXiv}},
  urldate = {2023-08-17},
  abstract = {There are two widely known issues with properly training Recurrent Neural Networks, the vanishing and the exploding gradient problems detailed in Bengio et al. (1994). In this paper we attempt to improve the understanding of the underlying issues by exploring these problems from an analytical, a geometric and a dynamical systems perspective. Our analysis is used to justify a simple yet effective solution. We propose a gradient norm clipping strategy to deal with exploding gradients and a soft constraint for the vanishing gradients problem. We validate empirically our hypothesis and proposed solutions in the experimental section.},
  archiveprefix = {arxiv},
  langid = {english},
  keywords = {Computer Science - Machine Learning},
  file = {/Users/lina/Zotero/storage/4UUXNARL/Pascanu et al. - 2013 - On the difficulty of training Recurrent Neural Net.pdf}
}
@misc{keras,
	title = {Keras: The high-level API for TensorFlow},
	url = {https://www.tensorflow.org/guide/keras},
	urldate={2023-08-18}
}
@misc{keras_tuner,
    title        = {KerasTuner},
    author       = {O'Malley, Tom and Bursztein, Elie and Long, James and Chollet, Fran\c{c}ois and Jin, Haifeng and Invernizzi, Luca and others},
    year         = 2019,
    howpublished = {\url{https://github.com/keras-team/keras-tuner}},
	urldate={2023-08-21}
}

@misc{simplilearnMachineLearningAlgorithms2018,
	title = {Machine {Learning} {Algorithms} {\textbar} {Machine} {Learning} {Tutorial} {\textbar} {Data} {Science} {Algorithms} {\textbar} {Simplilearn}},
	url = {https://www.youtube.com/watch?v=I7NrVwm3apg},
	urldate = {2023-04-25},
	author = {{Simplilearn}},
	month = mar,
	year = {2018},
	keywords = {ml},
}
@misc{mlp-vs-cnn-vs-rnn,
	title = {When to Use {MLP}, {CNN}, and {RNN} {Neural} {Networks}},
	author = {Brownlee, Jason},
	urldate = {2023-08-23},
	url={https://machinelearningmastery.com/when-to-use-mlp-cnn-and-rnn-neural-networks/}
}
@inproceedings{papandreaSmartphonebasedEnergyEfficient2012,
	title = {A smartphone-based energy efficient and intelligent multi-technology system for localization and movement prediction},
	doi = {10.1109/PerComW.2012.6197571},
	abstract = {Nowadays we are assisting to a noticeable proliferation of new generation smart-phones, as well as to the growth of mobile applications development. We are fairly surrounded by a huge number of proactive applications, which automatically provide users with relevant information exactly at the right time and at the right place when they need it. This ability is generally given by the exploitation of context information, and mainly Location Information. However, a location service running onto a smartphone has to deal with a great challenge, which is to manage the trade-off between the service's resources usage and its accuracy. In my work I investigate efficient localization strategies: these include, in addition to the standard location tracking techniques, the support of other technologies already available on mobile phones (i.e., sensors), as well as the integration of either Human Mobility Modelling and Machine Learning techniques. The main purposes of this work are: to reduce the impact that the service has on the device's resources usage in the case of continuous localization; to preserve the privacy of the user by running the whole system on the mobile device without relying on a back-end server; and to offer an ubiquitous coverage.},
	booktitle = {2012 {IEEE} {International} {Conference} on {Pervasive} {Computing} and {Communications} {Workshops}},
	author = {Papandrea, Michela},
	month = mar,
	year = {2012},
	keywords = {Accelerometers, Accuracy, Context, Mobile communication, Mobile handsets, Reliability, Sensors},
	pages = {554--555},
	file = {IEEE Xplore Abstract Record:/Users/lina/Zotero/storage/MMQDUP74/stamp.html:text/html;IEEE Xplore Full Text PDF:/Users/lina/Zotero/storage/WU6NWBRJ/Papandrea - 2012 - A smartphone-based energy efficient and intelligen.pdf:application/pdf},
}

@misc{IndoorNavigationComplete,
	title = {🧭 {Indoor} {Navigation}: {Complete} {Data} {Understanding}},
	shorttitle = {🧭 {Indoor} {Navigation}},
	url = {https://kaggle.com/code/andradaolteanu/indoor-navigation-complete-data-understanding},
	abstract = {Explore and run machine learning code with Kaggle Notebooks {\textbar} Using data from multiple data sources},
	language = {en},
	urldate = {2023-04-25},
	file = {Snapshot:/Users/lina/Zotero/storage/SWUMCBF2/indoor-navigation-complete-data-understanding.html:text/html},
}

@misc{DAWN2023,
	title = {{DAWN}},
	copyright = {GPL-2.0},
	url = {https://github.com/berlin-open-wireless-lab/DAWN},
	abstract = {Decentralized WiFi Controller},
	urldate = {2023-04-26},
	publisher = {Berlin Open Wireless Lab},
	month = apr,
	year = {2023},
	note = {original-date: 2017-05-24T13:38:03Z},
	keywords = {hacktoberfest},
}


@misc{LSTMKerasUnified,
	title = {{LSTM} by {Keras} with {Unified} {Wi}-{Fi} {Feats}},
	url = {https://kaggle.com/code/kokitanisaka/lstm-by-keras-with-unified-wi-fi-feats},
	abstract = {Explore and run machine learning code with Kaggle Notebooks {\textbar} Using data from multiple data sources},
	language = {en},
	urldate = {2023-04-26},
	file = {Snapshot:/Users/lina/Zotero/storage/AMYGE4W3/notebook.html:text/html},
}

@article{kaushikApproachDetectHuman2022,
	title = {An approach to detect human body movement using different channel models and machine learning techniques},
	volume = {13},
	issn = {1868-5145},
	url = {https://doi.org/10.1007/s12652-021-03237-2},
	doi = {10.1007/s12652-021-03237-2},
	abstract = {Worldwide, 16.7 million people die each year due to cardiovascular disease. These statistics raise demand of devices like sensor-based pacemakers (PM) which are not just doing heart rate augmentation but also capable to transmit information via wireless link to on body sensor and support remote monitoring of such patients. As per the world health organization WHO reports there are more than 3 million functioning PMs and about 600,000 pacemakers are implanted each year in world. On an average, 70–80\% of PMs are implanted in aged patients around 65 years or older. In addition to continuous monitoring of cardiovascular parameters, detection of physical movement of such patients may be helpful to assess their well-being. This paper has been formulated with an aim to highlight an approach which may be used to detect the physical movement of the patient using information signal received from implanted PM. The transmitted signal will experience a pathloss offered by wireless human body channel, which will affect the link quality parameters namely Signal to Noise Ratio (SNR) and Bit Error Rate (BER) and received signal strength indicator (RSSI). In the current work mathematical model has been formulated considering in body and on body channel propagation conditions and received power, received energy, pathloss, SNR, BER, bit rate, energy per bit and RSSI have been evaluated using IEEE802.15.6 channel models CM2 and CM3. Data set has been created and human body movement has been detected using Machine Learning (ML) techniques. Prediction accuracy of Multilayer Perceptron (MLP), k-Nearest Neighbours (kNN) and Random Forest have been compared. The analysis performed depicts that human body movement can be detected using different channel models and ML techniques such as MLP, kNN and Random Forest with an accuracy of 65.3\%, 72.8\% and 93.4\% respectively. The critical comparison of the result indicates that the performance of Random Forest is better than MLP and kNN. This approach will be helpful in remote detection of human body movement of patients.},
	language = {en},
	number = {8},
	urldate = {2023-04-26},
	journal = {Journal of Ambient Intelligence and Humanized Computing},
	author = {Kaushik, Monica and Gupta, Sindhu Hak and Balyan, Vipin},
	month = aug,
	year = {2022},
	keywords = {Bit error rate (BER), Random Forest, Received signal strength indicator (RSSI), Signal to noise ratio (SNR), Wireless Body Area Network (WBAN)},
	pages = {3973--3987},
	file = {Full Text PDF:/Users/lina/Zotero/storage/HG3EVJAD/Kaushik et al. - 2022 - An approach to detect human body movement using di.pdf:application/pdf},
}

@article{kimAPInitiatedFlowRedirection,
	title = {{AP}-{Initiated} {Flow} {Redirection} {Mechanism} for {AP} {Load} {Balancing} in {WLAN} {Environments}},
	abstract = {IEEE802.11 Wireless LAN (WLAN) is being widely used in public space such as airport, and increases the networking boundary in campus and enterprise, and it has lastly attracted considerable attention for mesh network and converged network with other 3G mobile communication networks. In WLAN, load balancing among Access Points (AP) is an important issue for efficient resource management or supporting the Quality of Service (QoS) of traffic, but most researches focused on the AP selection in network entry or roaming of Stations (STA). In this paper, we propose an AP-Initiated Flow Redirection (FR) for AP load balancing by monitoring AP’s availability in the true sense. When the AP’s resource becomes almost saturated, that is used more than a specific threshold, the AP queries the roaming possible neighbor APs about their availability and calculates the distribution of traffic load with statistical methods such as entropy or chi-square. Finally, the AP decides flows and new APs for redirection and performs it. Our simulation results show that our FR mechanism increases the performance in the various views.},
	language = {en},
	author = {Kim, Mihui and Chae, Kijoon},
	file = {Kim and Chae - AP-Initiated Flow Redirection Mechanism for AP Loa.pdf:/Users/lina/Zotero/storage/BYZBJISJ/Kim and Chae - AP-Initiated Flow Redirection Mechanism for AP Loa.pdf:application/pdf},
}

@article{yousefiSurveyBehaviorRecognition2017,
	title = {A {Survey} on {Behavior} {Recognition} {Using} {WiFi} {Channel} {State} {Information}},
	volume = {55},
	issn = {1558-1896},
	doi = {10.1109/MCOM.2017.1700082},
	abstract = {In this article, we present a survey of recent advances in passive human behavior recognition in indoor areas using the channel state information (CSI) of commercial WiFi systems. The movement of the human body parts cause changes in the wireless signal reflections, which result in variations in the CSI. By analyzing the data streams of CSIs for different activities and comparing them against stored models, human behavior can be recognized. This is done by extracting features from CSI data streams and using machine learning techniques to build models and classifiers. The techniques from the literature that are presented herein have great performance; however, instead of the machine learning techniques employed in these works, we propose to use deep learning techniques such as long-short term memory (LSTM) recurrent neural networking (RNN) and show the improved performance. We also discuss different challenges such as environment change, frame rate selection, and the multi-user scenario; and finally suggest possible directions for future work.},
	number = {10},
	journal = {IEEE Communications Magazine},
	author = {Yousefi, Siamak and Narui, Hirokazu and Dayal, Sankalp and Ermon, Stefano and Valaee, Shahrokh},
	month = oct,
	year = {2017},
	note = {Conference Name: IEEE Communications Magazine},
	keywords = {Antennas, Behavioral sciences, Doppler shift, OFDM, Receivers, Wireless communication, Wireless fidelity},
	pages = {98--104},
	file = {IEEE Xplore Abstract Record:/Users/lina/Zotero/storage/DC82KSGL/stamp.html:text/html;IEEE Xplore Full Text PDF:/Users/lina/Zotero/storage/WUVC7GJF/Yousefi et al. - 2017 - A Survey on Behavior Recognition Using WiFi Channe.pdf:application/pdf},
}

@article{szottWiFiMeetsML2022,
	title = {Wi-{Fi} {Meets} {ML}: {A} {Survey} on {Improving} {IEEE} 802.11 {Performance} {With} {Machine} {Learning}},
	volume = {24},
	issn = {1553-877X},
	shorttitle = {Wi-{Fi} {Meets} {ML}},
	doi = {10.1109/COMST.2022.3179242},
	abstract = {Wireless local area networks (WLANs) empowered by IEEE 802.11 (Wi-Fi) hold a dominant position in providing Internet access thanks to their freedom of deployment and configuration as well as the existence of affordable and highly interoperable devices. The Wi-Fi community is currently deploying Wi-Fi 6 and developing Wi-Fi 7, which will bring higher data rates, better multi-user and multi-AP support, and, most importantly, improved configuration flexibility. These technical innovations, including the plethora of configuration parameters, are making next-generation WLANs exceedingly complex as the dependencies between parameters and their joint optimization usually have a non-linear impact on network performance. The complexity is further increased in the case of dense deployments and coexistence in shared bands. 
				While classical optimization approaches fail in such conditions, machine learning (ML) is able to handle complexity. Much research has been published on using ML to improve Wi-Fi performance and solutions are slowly being adopted in existing deployments. In this survey, we adopt a structured approach to describe the various Wi-Fi areas where ML is applied. To this end, we analyze over 250 papers in the field, providing readers with an overview of the main trends. Based on this review, we identify specific open challenges and provide general future research directions.},
	number = {3},
	journal = {IEEE Communications Surveys \& Tutorials},
	author = {Szott, Szymon and Kosek-Szott, Katarzyna and Gawłowicz, Piotr and Gómez, Jorge Torres and Bellalta, Boris and Zubow, Anatolij and Dressler, Falko},
	year = {2022},
	note = {Conference Name: IEEE Communications Surveys \& Tutorials},
	keywords = {Machine learning, Wi-Fi, Wireless fidelity, 5G mobile communication, artificial intelligence, Artificial neural networks, deep learning, IEEE 802.11 Standard, IEEE 80211, machine learning, Radio frequency, Support vector machines, WLAN},
	pages = {1843--1893},
	file = {IEEE Xplore Full Text PDF:/Users/lina/Zotero/storage/YKIXX2LM/Szott et al. - 2022 - Wi-Fi Meets ML A Survey on Improving IEEE 802.11 .pdf:application/pdf},
}
@misc{neptune-ai,
	url = {https://neptune.ai/blog/select-model-for-time-series-prediction-task},
	urldate = {2023-08-23},
	title = {How to Select a Model For Your Time Series Prediction Task},
	author = {Joos Korstanje}	
}
@misc{xyz10technologyIndoorLocationCompetition2020,
	title = {Indoor {Location} {Competition} 2.0 {Webinar}},
	url = {https://www.youtube.com/watch?v=xt3OzMC-XMU},
	urldate = {2023-07-11},
	author = {{XYZ10 Technology}},
	month = aug,
	year = {2020},
}
@misc{multi-class-classification,
	url = {https://h2o.ai/wiki/multiclass-classification/},
	urldate = {2023-09-08},
	author = {H2O.ai},
	title = {What is Multiclass classification?}
}

@article{mlp_and_nn,
	title={Multilayer perceptron and neural networks},
	author={Popescu, Marius-Constantin and Balas, Valentina E and Perescu-Popescu, Liliana and Mastorakis, Nikos},
	journal={WSEAS Transactions on Circuits and Systems},
	volume={8},
	number={7},
	pages={579--588},
	year={2009},
	url = {https://www.academia.edu/download/69679997/29-485.pdf},
	publisher={World Scientific and Engineering Academy and Society (WSEAS) Stevens Point}
}

@article{seifertImplementingPositioningAlgorithms,
	title = {Implementing {Positioning} {Algorithms} {Using} {Accelerometers}},
	language = {en},
	author = {Seifert, Kurt and Camacho, Oscar},
	file = {Seifert and Camacho - Implementing Positioning Algorithms Using Accelero.pdf:/Users/lina/Zotero/storage/359AUWH3/Seifert and Camacho - Implementing Positioning Algorithms Using Accelero.pdf:application/pdf},
}

@book{BishopPatternRecognition,
	title = {{Pattern} {Recognition} and {Machine} {Learning}},
	author = {Christopher M. Bishop},
	year = {2006},
	isbn = {978-0387-31073-2},
	urldate = {2023-07-21},
	file = {Bishop - Pattern Recognition and Machine Learning.pdf:/Users/lina/Zotero/storage/HV93HWEF/viewer.html:text/html},
}

@misc{cat_cross_entropy,
	url = {https://gombru.github.io/2018/05/23/cross_entropy_loss/},
	author = {Raúl Gómez},
	urldate = {2023-08-31},
	title = {Understanding Categorical Cross-Entropy Loss, Binary Cross-Entropy Loss, Softmax Loss, Logistic Loss, Focal Loss and all those confusing names},
}

@misc{MulticlassClassificationNeurocomputing,
	title = {5. {Multi}-class classification — {Neurocomputing}},
	url = {https://julien-vitay.net/lecturenotes-neurocomputing/2-linear/5-Multiclassification.html},
	urldate = {2023-07-22},
	file = {5. Multi-class classification — Neurocomputing:/Users/lina/Zotero/storage/M5CWKF2G/5-Multiclassification.html:text/html},
}

@article{802.11k,
	title = {{IEEE} {Standard} for {Information} technology– {Local} and metropolitan area networks– {Specific} requirements– {Part} 11: {Wireless} {LAN} {Medium} {Access} {Control} ({MAC})and {Physical} {Layer} ({PHY}) {Specifications} {Amendment} 1: {Radio} {Resource} {Measurement} of {Wireless} {LANs}},
	shorttitle = {{IEEE} {Standard} for {Information} technology– {Local} and metropolitan area networks– {Specific} requirements– {Part} 11},
	doi = {10.1109/IEEESTD.2008.4544755},
	abstract = {This amendment specifies the extensions to IEEE Std 802.11 for Wireless LANs providing mechanisms for Radio Resource Measurement.},
	journal = {IEEE Std 802.11k-2008 (Amendment to IEEE Std 802.11-2007)},
	month = jun,
	year = {2008},
	note = {Conference Name: IEEE Std 802.11k-2008 (Amendment to IEEE Std 802.11-2007)},
	keywords = {802.11k-2008, IEEE Standards, Information exchange, Information technology, local area network (LAN), measurement, network management, radio, radio resource, Telecommunications, Wireless LAN},
	pages = {1--244},
	file = {IEEE Xplore Full Text PDF:/Users/lina/Zotero/storage/IA5HP9H7/2008 - IEEE Standard for Information technology– Local an.pdf:application/pdf},
}

@article{IEEEStandardInformation2008a,
	title = {{IEEE} {Standard} for {Information} technology– {Local} and metropolitan area networks– {Specific} requirements– {Part} 11: {Wireless} {LAN} {Medium} {Access} {Control} ({MAC})and {Physical} {Layer} ({PHY}) {Specifications} {Amendment} 1: {Radio} {Resource} {Measurement} of {Wireless} {LANs}},
	shorttitle = {{IEEE} {Standard} for {Information} technology– {Local} and metropolitan area networks– {Specific} requirements– {Part} 11},
	doi = {10.1109/IEEESTD.2008.4544755},
	abstract = {This amendment specifies the extensions to IEEE Std 802.11 for Wireless LANs providing mechanisms for Radio Resource Measurement.},
	journal = {IEEE Std 802.11k-2008 (Amendment to IEEE Std 802.11-2007)},
	month = jun,
	year = {2008},
	note = {Conference Name: IEEE Std 802.11k-2008 (Amendment to IEEE Std 802.11-2007)},
	keywords = {802.11k-2008, IEEE Standards, Information exchange, Information technology, local area network (LAN), measurement, network management, radio, radio resource, Telecommunications, Wireless LAN},
	pages = {1--244},
	file = {IEEE Xplore Abstract Record:/Users/lina/Zotero/storage/72MLARJD/4544755.html:text/html;IEEE Xplore Full Text PDF:/Users/lina/Zotero/storage/Z9XL8BM2/2008 - IEEE Standard for Information technology– Local an.pdf:application/pdf},
}

@article{802.11r,
	title = {{IEEE} {Standard} for {Information} technology– {Local} and metropolitan area networks– {Specific} requirements– {Part} 11: {Wireless} {LAN} {Medium} {Access} {Control} ({MAC}) and {Physical} {Layer} ({PHY}) {Specifications} {Amendment} 2: {Fast} {Basic} {Service} {Set} ({BSS}) {Transition}},
	shorttitle = {{IEEE} {Standard} for {Information} technology– {Local} and metropolitan area networks– {Specific} requirements– {Part} 11},
	doi = {10.1109/IEEESTD.2008.4573292},
	abstract = {This amendment specifies the extensions to IEEE Std 802.11-2007 for wireless local area networks (WLANs) providing mechanisms for fast basic service set (BSS) transition.},
	journal = {IEEE Std 802.11r-2008 (Amendment to IEEE Std 802.11-2007 as amended by IEEE Std 802.11k-2008)},
	month = jul,
	year = {2008},
	note = {Conference Name: IEEE Std 802.11r-2008 (Amendment to IEEE Std 802.11-2007 as amended by IEEE Std 802.11k-2008)},
	keywords = {WLAN, IEEE Standards, Information exchange, Information technology, Telecommunications, Wireless LAN, 802.11r-2008, LAN, local area network, wireless LAN},
	pages = {1--126},
	file = {IEEE Xplore Abstract Record:/Users/lina/Zotero/storage/UPH3F8S2/4573292.html:text/html;IEEE Xplore Full Text PDF:/Users/lina/Zotero/storage/RHKIATW5/2008 - IEEE Standard for Information technology– Local an.pdf:application/pdf},
}
@article{cisco-802.11,
	title = {802.11r Fast Transition Roaming},
	urldate = {2023-08-22},
	url = {https://www.cisco.com/c/en/us/td/docs/wireless/controller/technotes/5700/software/release/ios_xe_33/11rkw_DeploymentGuide/b_802point11rkw_deployment_guide_cisco_ios_xe_release33/b_802point11rkw_deployment_guide_cisco_ios_xe_release33_chapter_01.html}
}