Skip to content

Latest commit

 

History

History
17 lines (13 loc) · 779 Bytes

File metadata and controls

17 lines (13 loc) · 779 Bytes

Analysing Amazon Product Reviews using Dimensions of Valence and Arousal in Python

Techniques:

OpenAI/GPT3, Random Forest Classification, SVM, Fast Clustering

Libraries Used:

  1. Loading, exploring and manipulating datasets: pandas, numpy, random, google.colab
  2. Error Handling for API calling and importing API keys from envirnoment: retry, time, os
  3. Using gpt3 completions, embeddings and generating multiclass precion-recall: openai
  4. For tokeizing text to perform validation for GPT3: transformers
  5. ML models for classifiers, dimensionality reduction, model analysis: sklearn
  6. For fast clustering and embeddings: sentence_transformers, torch
  7. Visualisation: matplotlib
  8. Model exports & imports: import pickle
  9. for removing warnings: import warnings