The objective of this work was to develop machine learning models (Support Vector Machines, Random Forest, Gradient Boosted Trees) that could accurately predict adverse drug reactions using the SIDER and OFFSIDES databases and Python (pandas, numpy, scikit-learn, matplotlib, xgboost, rdkit, imblearn).
The notebook can be found here GitHub Notebook or here nbviewer. The code can be found here GitHub Code.
Position: 109th/1214, Top 9%
The goal of this competition was to provide a simple extension to the classic MNIST competition. Instead of using Arabic numerals, it used a recently-released dataset of Kannada digits.
I used a Convolutional Neural Networks, optimized it, used data augmentation, among other techiniques. The final results were a private score of 0.9896 and a public score of 0.9886 (accuracy).
The notebook can be found here Kaggle Notebook.
This work consists in a EDA and a Regression Task using a kaggle dataset containing information about restaurants in Bangalore. The main objective of this analysis was to understand which factors could affect the rate of a restaurant and if this can be predicted within a small margin of error.
The notebook can be found here GitHub Notebook or here nbviewer. The kaggle page can be found here.
A quick EDA and statistical analysis in a kaggle dataset contaning marks secured by the students in various subjects. The main objective of this analysis was to understand which factors could affect these grades.
The notebook can be found here GitHub Notebook or here nbviewer. The kaggle page can be found here.
Code and figures for my #TidyTuesday projects, a weekly data project aimed at the R ecosystem.