ecancellieri

Web_Traffic_Time_Series_Forecasting

Based on the Kaggle competition "Web Traffic Time Series Forecasting (https://www.kaggle.com/c/web-traffic-time-series-forecasting) It uses an ARIMA model to predict pages visualisation. It use the key_2.csv file from the Kaggle competition page with the list of pages and dates for each page to predict, and use the train_2.csv file as training dataset

Multi_Armed_Bandit

Solution of the multi-armed bandit problem following R. S. Sutton and A. G. Barto

Out_of_Equilibrium_Nonlinear_SSH

This code evaluates the time evolution of a set of differential equations correspoding to an out of equilibrium Su-Schrieffer-Heeger (SSH) chain. It includes saturable pumps on first and last sites, and linear decays on all sites. It includes three different numerical solvers.

demographics

demographics

Bike_Sharing_Deman_With_bsts

From the London Data Science Workshop a quick test with the bsts library in r

CodeWars

My exercises on CodeWars

Face_Recognition_Koibumi

first-timers

Gender_Bias

MLOpsPython_EC

ecancellieri/named_entity_recognition

Testing how to build a named entity recognition model

ecancellieri/MLOpsPython_EC

ecancellieri/Gender_Bias

ecancellieri/Face_Recognition_Koibumi

ecancellieri/simple-maths

Simple math functions in (mostly) python

ecancellieri/My_twitter_apps

I place here some projects based on Twitter data

ecancellieri/first-timers

ecancellieri/Out_of_Equilibrium_Nonlinear_SSH

This code evaluates the time evolution of a set of differential equations correspoding to an out of equilibrium Su-Schrieffer-Heeger (SSH) chain. It includes saturable pumps on first and last sites, and linear decays on all sites. It includes three different numerical solvers.

ecancellieri/CodeWars

My exercises on CodeWars

ecancellieri/Bike_Sharing_Deman_With_bsts

From the London Data Science Workshop a quick test with the bsts library in r

ecancellieri/Quant_Agent_Sinusoidal

A reinforcement learning code for an agent that learns to trade. The market is a simple sinusoidal curve, the Q function is a Neural Network with 2 layers, the possible actions are either to buy (1) or to hold(0)

ecancellieri/Web_Traffic_Time_Series_Forecasting

Based on the Kaggle competition "Web Traffic Time Series Forecasting (https://www.kaggle.com/c/web-traffic-time-series-forecasting) It uses an ARIMA model to predict pages visualisation. It use the key_2.csv file from the Kaggle competition page with the list of pages and dates for each page to predict, and use the train_2.csv file as training dataset

ecancellieri/Spooky_Authors

Code to participate to the Kaggle training competition to recognise authors from texts

ecancellieri/Multi_Armed_Bandit

Solution of the multi-armed bandit problem following R. S. Sutton and A. G. Barto