This work attempts to generalize a stock forecasting neural network using Bayesian regularization so that predictions can be performed without an overfitted model, considering the highly volatile market these days.

Classification Using Logistic Regression by Making a Neural Network Model. This project also includes comparison of Model performance when different regularization techniques are used

Machine Learning and Data Mining Projects (2022-2023)

Regularized maximum likelihood estimation for discrete choice models on the LPMC dataset

the implementation of a multilayer perceptron

Constructing a linear regression model which may help us predict the happiness score in the 155 countries.

Code for the paper "Module-based regularization improves Gaussian graphical models when observing noisy data"

Analyzing car accident fatalities to pave the way for preventative measures and safer transportation using Statistical and Machine Learning algorithms

Digital Image Reconstruction

A school bootcamp for hands on learning of Machine Learning

Conducted advanced research on unsupervised image segmentation using Lipschitz regularity constraints

The objective is to build various classification models, tune them and find the best one that will help identify failures so that the generator could be repaired before failing/breaking and the overall maintenance cost of the generators can be brought down.

In linear regression, regularization is a process of making the model more regular or simpler by shrinking the model coefficient to be closer to zero or absolute, ultimately to address over fitting.

Functional Group Bridge for Simultaneous Regression and Support Estimation (https://arxiv.org/abs/2006.10163)

Repository with some implementations of algorithms used in Numerical Analysis. From the solution of determined and overdeterminded systems to regularization and non-linear least square problems.

This repository contains a Python implementation of linear regression, logistic regression, and ridge regression algorithms. These algorithms are commonly used in machine learning and statistical modeling for various tasks such as predicting numerical values, classifying data into categories, and handling multicollinearity in regression models.

Using deep learning to predict whether students can correctly answer diagnostic questions

Supervised learning and unsupervised in R, with a focus on regression and classification methods.

Preprocessed the USPS dataset, implemented and compared different network architectures and optimization techniques, applied regularization techniques such as ensembling and dropout, performed adversarial training to evaluate network robustness, and evaluated network performance using metrics such as accuracy, precision, and recall.