Sugata Ghosh's repositories
E-commerce-Text-Classification
Proper categorization of e-commerce products enhances the user experience and achieves better results with external search engines. The objective of the project is to classify a product into four given categories, based on its description available on an e-commerce platform.
Electron-Energy-Flux-Prediction-using-Deep-Learning
The objective of the project is to predict the total electron energy flux based on various relevant features, in the context of modeling electron particle precipitation from the magnetosphere to the ionosphere.
Implementing-Logistic-Regression-from-Scratch
While it is convenient to use advanced libraries for day-to-day modeling, it does not give insight into the details of what really happens underneath, when we run the codes. In this work, we implement a logistic regression model manually from scratch, without using any advanced library, to understand how it works.
Anomaly-Detection-in-Credit-Card-Transactions
The objective of the project is to detect anomalies in credit card transactions. More precisely, given the data on time, amount and 28 transformed features, our goal is to fit a probability distribution based on authentic transactions, and then use it to correctly identify a new transaction as authentic or fraudulent.
Higgs-Boson-Event-Detection
The goal of the project is to classify an event produced in the particle accelerator as background or signal. A background event is explained by the existing theories and previous observations. A signal event, however, indicates a process that cannot be described by previous observations and leads to the potential discovery of a new particle.
Site-Energy-Usage-Intensity-Prediction
The objective of the project is to predict energy usage intensity of a building in a given year, based on building characteristics as well as weather data for the location of the building. The dataset contains about 100 thousand observations of building energy usage records collected over 7 years, in several states within the United States.
SpaceX-Falcon-9-First-Stage-Landing-Prediction
A primary reason behind SpaceX's stellar success is its relatively inexpensive rocket launches, founded on its ingenious reuse of the first stage of a rocket. Thus, to estimate the cost of a launch, it is critical to predict if the first stage will land. In this project, we attempt to predict first stage landing of SpaceX's Falcon 9 rocket.
Road-Traffic-Accident-Severity-Classification
The aim of the project is to build prediction models to classify severity of road traffic accidents (slight injury, serious injury or fatal injury) based on various relevant information regarding the involved vehicles, drivers, casualties and surrounding conditions.
Spooky-Author-Identification
The objective of the project is to train an LSTM model with the help of GloVe embeddings, to predict probabilities that a given text is written by particular authors. Furthermore, these probabilities are used to predict the author of the text in a multiclass classification setup.
Credit-Card-Fraud-Detection
The objective of the project is to classify credit card transactions as authentic or fraudulent, based on data regarding time, amount and a set of PCA-transformed features for transactions. We explore the data extensively and employ different techniques to build classification models, which are compared through various evaluation metrics.
Foursquare-Location-Matching
Using the provided dataset of over one-and-a-half million place entries, heavily altered to include noise, duplications, extraneous, or incorrect information, the objective is to produce an algorithm that predicts which place entries represent the same point of interest (POI).
Natural-Language-Processing-with-Disaster-Tweets
The objective of the project is to predict whether a particular tweet, of which the text (occasionally the keyword and the location as well) is provided, indicates a real disaster or not. We use various NLP techniques and classification models for this purpose and objectively compare these models by means of appropriate evaluation metric.
dsml
Data Science and Machine Learning Portfolio Website
Patient-Survival-Prediction-using-Deep-Learning
The knowledge about medical records is crucial in the treatment of a patient and it indicates his/her survival odds to a great extent. In this project, we predict whether a patient will survive or not, based on various relevant medical information.