Semester 2

PROJECTS

Title: Fake Income Allocation And  Classification

Synopsis: Subsidy Income is a company who delivers subsidies to individuals based on their income. But now a days, it’s very much difficult to know the actual income of every individual. Due to the non-availability of accurate data on financial parameters, incompleteness in regard to the coverage is embedded in such the dataset. Subsidy Income company wishes to develop an income classification-system for individuals using some statistical methods to enhance the accuracy-level of the data and to create a good classification using some high performance algorithms, like, Logistic Regression algorithm ,K-Nearest Neighbours Classifier algorithm and Decision Tree algorithm using python software.

Student Name :

-----------------------------------

Supervisor: Mr. Prasenjit Banerjee

Title: Unmasking Real and Fraudulent Job Classifier Model by Machine Learning and NLP

Synopsis:In the current job market, where employment opportunities are scarce, individuals seeking jobs often encounter fraudulent job postings and scams. These deceptive job listings not only waste the time and effort of job seekers but also pose significant risks to their personal information and financial security. To address this issue, machine learning models and natural language processing (NLP) techniques are utilized to predict the likelihood of a job posting being genuine or fraudulent.The objective of this project is to compare five classification models, namely the Logistic Regression Model, Random Forest Model, Multinomial Naïve Bayes Model, SGD Classifier Model, and XGBoost, to identify an effective model that can accurately classify fake and real job postings. Several factors such as job description, location, salary range, and company information are considered in this classification process. The project aims to achieve high precision in distinguishing between legitimate and fraudulent job postings. By doing so, it strives to contribute to the creation of a safer and more trustworthy job market for job seekers worldwide .

Student Name :

-----------------------------------

Supervisor: Mr. Chandan Chakraborty

Title: A Study of India’s employment condition using Time Series Analysis

Synopsis: India has been facing the issue of unemployment, with the rate reaching its highest in the last four decades. The COVID- 19 pandemic has had a significant impact on the job market, leading to job losses across various sectors, particularly in the informal and unorganized sectors. The government has launched various initiatives to address the issue of unemployment, such as the National Employment Policy and the Atmanirbhar Bharat Abhiyan, but sustained efforts are needed to create more job opportunities for citizens. In this paper It is has been predicted that what will be the unemployment rate of India and West Bengal at the end of financial year 2024. The model which is used in this paper is called VARMA (Vector Auto Regressive Moving Average) model.

Student Name :

-----------------------------------

Supervisor: Dr. Sushovon Jana

Title: Study on Network Analysis in Financial Market : Centrality Measure and Beta Factor

Synopsis: Financial markets are complex systems where interconnections and interdependence between various market participants play a crucial role in determining market dynamics. This introduction presents a study focused on network dynamics in financial markets, specifically analyzing centrality measures and beta factors. The analysis of centrality measures provides insights into the importance of individual market participants .In addition to the beta factor, we will also examine the centrality measures in the Indian financial market network. Centrality measures, such as eigenvector centrality , harmonic centrality , pagerank , closeness centrality, and betweenness centrality, provide insights into the importance and influence of individual nodes or market participants within the network.

Student Name :

-----------------------------------

Supervisor: Ms. Anwesha Sengupta & Prof. Prasanta Narayan Dutta

Title: A Study on Principal Component Analysis and Regression on Air Pollution Data

Synopsis: In this project we are trying to compare Multiple Linear Regression, Principal Component Regression. Generally to predict a response we apply MLR and we check the value of R-Squared to see if the model is a good fit. But we wanted to check if PCR gives better value of R-Squared.We also wanted to check the in a particular city which pollutants are causing high AQI.

Student Name :

-----------------------------------

Supervisor: Mr. Anjan Samanta

Title: Diagnosis of Polycystic Ovarian Syndrome and Identification of Key Features associated with it Using Machine Learnings Algorithms

Synopsis: The objective of this research is to identify the most important parameters for diagnosing PCOS using Supervised Feature Selection Algorithms. The analysis aims to determine whether PCOS can be identified using a reduced number of parameters. This information can assist doctors in diagnosing PCOS accurately and enable researchers to save time by eliminating irrelevant parameters in future PCOS-related studies. Additionally, machine learning classifiers will suggest the most appropriate ML model to be implemented by diagnostic centers, ensuring optimal sensitivity and specificity during PCOS diagnosis.

Student Name :

-----------------------------------

Supervisor: Mr. Taranga Mukherjee & Dr. Indrani Mukherjee 

Title: Prediction of forest fire's occurrence 

Synopsis: Accurate prediction of forest fire can help reducing the damages and loss.

Student Name :

-----------------------------------

Supervisor: Mr. Debjit Konai