We will discuss the Hyper Parameter Tuning for different Machine Learning Algorithm

Classification model to predict the probability that a customer defaults based on their monthly customer statements using the data provided by American Express.

Hyperparameter optimization or tuning is the problem of choosing a set of optimal hyperparameters for a learning algorithm. A hyperparameter is a parameter whose value is used to control the learning process. By contrast, the values of other parameters (typically node weights) are learned. Hyperparameters are crucial as they control the overall…

RandomSearch CV vs Grid Search

A Streamlit web app utilizing Python, scikit-learn, and pandas for used car price prediction. Features data preprocessing (scaling, encoding), Random Forest model optimization with GridSearchCV, and interactive user input handling. Achieves high accuracy (R² score: 0.9028), showcasing skills in machine learning, data engineering, and deployment.

Advanced ML Case Study where we use ML algorithms to detect malware from a given piece of software.

Enhancing The Performance Of Classifiers In Detecting Abnormalities In Medical Data Using Nature Inspired Optimization Techniques

Analyzing a dataset of bank transactions and using gradient boosting classifier to capture as many fraudulent transactions as possible while minimizing false positives.

Algorithms used to confirm whether a celestial body is a planet or not.

A machine learning project to predict loan defaults in a German bank's customer base. Using the German Credit Risk dataset, it explores key factors contributing to defaults and trains models like Random Forest, GBM, and XGBoost. Includes EDA, data processing, hyperparameter tuning, and model evaluation.

This is a Premiere Project done by Team Gitlab in Hamoye Data Science Program Dec'22. Out of 5 models used on the data, Random Forest Classifier was used to further improve the prediction of characters death. With parameter tuning and few cross validation, we were able to reduce the base error by 5.42% and increase accuracy by 2,42%.

🌟 Time Series Forecasting for Industrial Wastewater - Predicting heavy metal concentrations using advanced models like **ARIMA** and **PSO-LSTM**, blending statistical and machine learning techniques to enhance wastewater treatment efficiency. 🚀

Predict precipitation to mitigate flood damage in Bangladesh

This project aims to develop a machine learning model to predict bike-sharing demand based on various factors such as weather conditions, time of day, and historical usage patterns. The dataset used for this project consists of 8760 records and 14 attributes.

Implementation of Hyper-parameter tuning of ML models

Credit score prediction using classification models (Multi-class prediction)

This is an End-to-End Data Science Project built in order to help an International E-commerce Company to predict whether their product will be delivered on the committed Delivery Time or not

GridSearchCV, RandomSearchCV For Model optimization and Saving/Loading the model

List of completed academic projects

Machine Learning with Sklearn