Marc Aurele Divernois's Projects
AB Testing of conversion rates
Bayesian Optimization for hyperparameter choice
Convolution Neural Networks for digit classification
Learn word embeddings on Harry Potter corpus
Gaussian mixture model for predicting customer churn
K-means for fraud/anomaly detection
Naive Bayes classifier for spam detection
Principal Component Analysis theory and use case on toy data
Random forest on Kaggle use case
Movie recommender system based on IMDB data
SVM theory and use case on toy data
MATLAB codes to replicate Duan et al. 's paper "Multiperiod corporate default prediction—A forward intensity approach"
Source code, data and plots for our paper "Analysis of Large Market Data Using Neural Networks: A Causal Approach"
Source codes and plots for my paper "A Deep Learning Approach to Estimate Forward Default Intensities"
Kaggle competition - Titanic
An implementation of a complete machine learning solution in Python on a real-world dataset. This project is meant to demonstrate how all the steps of a machine learning pipeline come together to solve a problem!
Learn how to responsibly deliver value with ML.
Credit card fraud detection with unlabeled data. Algos : PCA, K-means, ANN, Logistic
K-Means Clustering to Identify Handwritten Digits
LaTeX codes for my PhD thesis at EPFL
input-app-svm-classification
Word Mini-Game : Guess the secret word ! Play here :
General purpose unsupervised sentence representations
Source codes to scrape tweets from the Stocktwits API and store as JSON. Preprocessing steps for NLP classification.
Digit classification using SVM
Deep learning library featuring a higher-level API for TensorFlow.
Contains different algos for timeseries analysis
NLP repo for our paper "StockTwits Classified Sentiment and Stock Returns". StockTwits data is available in the README.