Aditya Choudhury

A 3rd year bachelor's student with logical, analytical skills, interest in the field of Software development, Machine Learning and evolving Technologies. Self-motivated with focus on Team building and Leadership to bring synergy at workplace. Good at Communication and designing. Inquisitive, energetic with strong foundation in math and logic.

• WEB DEVELOPMENT, Acmegrade, 02/2024
• CLOUD COMPUTING, Internforte, 02/2024
• PROBLEM SOLVING, HackerRank, 06/2023
• CLOUD COMPUTING, AWS, 07/2023

SALARY PREDICTION (03/2024)

A straightforward machine learning model designed for salary prediction

• Conducted a comprehensive study on salary prediction utilizing a Python notebook environment.
• Utilized popular libraries such as Scikit-Learn and Pandas for data preprocessing, analysis, and model development.
• Employed Multiple Linear Regression to construct a predictive model for estimating salary based on variables like years of experience, current position, and job location.
• Applied advanced data analysis techniques to identify key features and enhance model performance.
• Conducted thorough testing and evaluation to validate the accuracy and reliability of the salary prediction model.
• Provided valuable insights to enhance understanding of factors impacting salary discrepancies in the job market.

DRUG CLASSIFICATION (04/2024)

Drug Classification Neural Network

• Designed and implemented a linear feed-forward neural network for drug classification tasks
• Incorporated the sigmoid activation function within the neural network architecture
• Trained the model to predict the administered drug based on basic electrolytic variables
• Classified patients into five drug categories: DrugA, DrugB, DrugC, DrugD, and DrugE
• Applied backpropagation for error correction to optimize model performance
• Utilized the Adam optimizer for efficient optimization during training

HOUSE PRICE PREDICTION MODEL (04/2024)

House Price Prediction Using Random Forest Regression 

• Developed a predictive model using random forest regression to estimate house prices based on key features.
• Gathered and cleaned housing dataset, ensuring data quality and integrity for analysis.
• Implemented random forest regression algorithm in Python, utilizing scikit-learn library for model training and evaluation.
• Tuned hyperparameters to optimize model performance and improve prediction accuracy.
• Evaluated model performance using metrics such as mean absolute error (MAE) and root mean squared error (RMSE).
• Successfully predicted house prices with an accuracy of 97%, demonstrating proficiency in machine learning techniques.
• Enhanced skills in data preprocessing, feature engineering, and model evaluation through practical application.

SPAM MAIL DETECTION (02/2024)

Emails undergo classification into two categories, namely 'Ham' and 'Spam,' employing machine learning algorithms. This approach operates under the assumption that word occurrences are independent of each other.

• Developed and implemented a spam mail detection system utilizing machine learning algorithms
• Utilized NLTK (Natural Language Toolkit) for pre-processing techniques to filter out stop-words and extract meaningful words from email data
• Employed count vectorization to convert pre-processed emails into numerical data for machine learning algorithms
• Trained the model using the Multinomial Naive Bayes classification algorithm for effective text classification
• Applied word independence assumption to create a robust classifier for accurate classification of emails into 'Ham' (non-spam) and 'Spam' categories
• Conducted extensive testing and evaluation to ensure the reliability and efficiency of the detection system
• Successfully deployed the system for real-world use, contributing to the reduction of spam emails and enhancing overall email security.