Sayan De's Projects
This repository serves as a comprehensive guide to deep learning concepts, designed to evolve from fundamental ideas to advanced techniques. Starting with the basics of perceptrons and moving through the intricacies of multilayer perceptrons (MLPs), this repository aims to provide a structured learning path for anyone interested in Deep learning
Developing a robust deep learning model for breast cancer classification, leveraging ML techniques to differentiate malignant and benign tumors from tissue data. Using CNNs and diverse datasets, we aim to enhance medical diagnostics, aiding informed healthcare decisions and improving patient outcomes
Developed a credit risk analysis project using Keras and TensorFlow. Implemented neural networks to assess and predict credit risk based on historical financial data. Features include model training, evaluation, and predictions to help financial institutions manage risk and make informed lending decisions.
create a user-friendly web app using Streamlit, predicting diabetes risk from health data. Machine learning models trained on medical features like glucose, blood pressure, BMI offer personalized predictions. We empower users to make informed health decisions, potentially preventing diabetes.
Utilizing MobileNet transfer learning in TensorFlow, this project classifies cat and dog images. With Kaggle data, it fine-tunes pre-existing weights, integrates OpenCV for preprocessing, and PIL for augmentation. The aim is to optimize model performance in distinguishing between the two, advancing computer vision for pet classification
The project aims to perform Exploratory Data Analysis (EDA) on two internal datasets provided by the Seaborn library in Python: the Iris dataset and the Flights dataset. EDA is a crucial step in any data analysis process, allowing us to gain insights, identify patterns, and understand the main characteristics of the data.
Leveraging ColumnTransformer, pipelines, standardization, and encoding, we'll preprocess data. Using Logistic Regression, Decision Trees, Random Forest, and XGBoost, we'll analyze factors like job satisfaction, promotion, and salary to predict churn. This helps companies improve satisfaction, reduce turnover, and enhance stability.
The project entails conducting Exploratory Data Analysis (EDA) on employee demographic data using Python. Through data visualization and statistical analysis, the project aims to uncover insights into various demographic factors such as age, gender, ethnicity, education, and tenure within the workforce. By exploring patterns, distributions, and cor
Crafting a comprehensive Excel dashboard for employee salary comparison, utilizing Pivot Tables/Charts with interactive filters (date, gender, ethnicity, business unit, department) for thorough analysis and exploration.
This project explores Bagging (Bootstrap Aggregating) for both classification and regression tasks. By training multiple models on different data subsets, Bagging improves accuracy and resilience. Using synthetic and real-world datasets, the project demonstrates how Bagging enhances performance over single-model approaches.
This project investigates ensemble learning techniques, combining multiple models to enhance accuracy and robustness. It covers both basic methods (Max Voting, Averaging, Weighted Averaging) and advanced techniques (Stacking, Blending, Bagging, Boosting), aiming to improve predictive performance by addressing model weaknesses.
Title: Excel Data Operations: Clean, Optimize, Analyze Description: Dive into Excel operations covering data cleaning, optimization, and analysis. Learn key functions like aggregates, pivot tables, VLOOKUP, macros, and charts for comprehensive data manipulation and visualization.
Analyze Hotel booking dataset to address a specific problem. Clean, explore, optimize data with Pandas, Matplotlib, Seaborn. Unearth trends, forecast booking/cancellation behaviors tied to pricing, country, market segments. Visualize insights using bar, pie, line charts.
Utilize OpenCV & face_recognition library to enable webcam-based face recognition. Encode and compare facial features in real-time for accurate identification. Tutorial covers environment setup, webcam video capture, face detection, encoding, comparison, and live recognition.
This project evaluates logistic regression, random forest, decision tree, and gradient boosting classifier models for fake news detection. Using labeled data, it analyzes accuracy, confusion matrices, and ROC curves to understand each model's effectiveness in discerning between real and fake news.
In this repository i will be storing all the Jupyter notebook ipynb files and dataset files csv and excel that are used for Exploratory data analysis, statistical analysis, Machine learning etc in the Analytics class
Forecasting gold prices with machine learning, employing Linear Regression and Naive models. Analyzing historical data to predict future prices, aiding decision-making in financial markets.
This project leverages TensorFlow and Keras to implement deep learning techniques, focusing on CNNs for recognizing handwritten digits from images. Integrated with OpenCV, it ensures precise digit extraction through robust image preprocessing and manipulation.
Developed a heart disease prediction model using Decision Tree algorithms. Analyzed patient data to classify risk levels and predict the likelihood of heart disease. Features include data preprocessing, model training, and performance evaluation, aimed at improving early diagnosis and personalized healthcare.