kanban-plugin |
---|
board |
- Create final presentation
- Create MLOps Part for Scientific Report #Deployment
- Create Model Validation Part for Scientific Report #Evaluation
- Create Methods we used for decision makings and include them into the Scientific Report #DataCollecting #DataPreparation #Modeling
- Create Data Processing Part for Scientific Report #DataPreparation
- Plan Cloud-Service task and compare it with all services, also with costs #BusinessUnderstanding
- Create Literature review for Scientific Report #BusinessUnderstanding #DataPreparation #Modeling
- Deploy training on VastAI #Modeling
- Try Transformer also to Resnet #Modeling
- Play the game and make a baseline accuracy for amateur {Yutaro, Lukas}, and pro {Linus} #Modeling
- Focus just on predicting accuracy of guessing countries #BusinessUnderstanding
- Deploy Training Iteration on other cloud service than gpuhub #Modeling
- Add Multiple labels architecture, combine coordinates with countries #Modeling
- Research other metrics for the evaluating the model #Modeling #Evaluation
- Add test metrics to the code #Evaluation
- Try more different base line models, Transformer Encoders, more Resnet architechtures and Vanilla CNN #Modeling
- Try other models like MobileNetV2 and a CNN too? #Modeling
- Change accuracy for best, 3 best and 5 best predictions, and log it, when it is under the predictions it's correct #Modeling
- Setup wand.db Experiment tracking for Model Validation #Evaluation
- Setup wand.db Experiment tracking for Model Training #Modeling
- Do some research about CNN or other networks which can handle images #Modeling
- up-to-date kanban board before each coaching session #WorkingPermanently
- Update Scientific Report for DataProcessing, Methods, Model Validation, MLOps #WorkingPermanently
- Focus on the task "Country Prediction" #WorkingPermanently
- create web scraper for geoguessr #DataCollecting
- collecting data from geoguessr #DataCollecting
- create commitment and proposal as markdown file #BusinessUnderstanding
- Create midterm presentation
- Create Proposal and upload it to Ilias #BusinessUnderstanding
- decide and vote for best business case title #BusinessUnderstanding
- Resize images to usable datatype and file sizes #DataPreparation
- Collect 60k images from Geoguessr #DataCollecting
- Add sample dataset to GitLab for Data Processing #DataCollecting
- Map coordinates to countries #DataPreparation
- Map coordinates to real coordinates #DataPreparation
- Get Data insights from images #DataUnderstanding
- Create a structure for data science project #DataPreparation #DataUnderstanding #Modeling #Evaluation #Deployment
- Create CNN-Architecture for the Model #Modeling
- Define base metric for training and evaluating #Modeling
- Load pre-trained weights and do some transfer learning , freeze some layers and unfreeze dense layers #Modeling
- Create Baseline Model #Modeling
%% kanban:settings
{"kanban-plugin":"board","date-colors":[{"isToday":false,"distance":1,"unit":"days","direction":"after"}],"tag-colors":[{"tagKey":"#BusinessUnderstanding","color":"rgba(255, 255, 255, 1)","backgroundColor":"rgba(255, 247, 0, 0.75)"},{"tagKey":"#DataCollecting","color":"rgba(255, 255, 255, 1)","backgroundColor":"rgba(72, 105, 34, 1)"},{"tagKey":"#DataUnderstanding","color":"rgba(255, 255, 255, 1)","backgroundColor":"rgba(12, 60, 22, 1)"},{"tagKey":"#DataPreparation","color":"rgba(255, 255, 255, 1)","backgroundColor":"rgba(25, 148, 139, 1)"},{"tagKey":"#Modeling","color":"rgba(255, 255, 255, 1)","backgroundColor":"rgba(50, 94, 171, 1)"},{"tagKey":"#Evaluation","color":"rgba(255, 255, 255, 1)","backgroundColor":"rgba(51, 39, 130, 1)"},{"tagKey":"#Deployment","color":"rgba(255, 255, 255, 1)","backgroundColor":"rgba(115, 55, 133, 1)"},{"tagKey":"#BugDetected","color":"rgba(255, 255, 255, 1)","backgroundColor":"rgba(255, 0, 0, 0.66)"},{"tagKey":"#WorkingPermanently","color":"rgba(255, 255, 255, 1)","backgroundColor":"rgba(196, 113, 22, 0.79)"}]}
%%