🧠 Data Science & Machine Learning Portfolio

📁 Repository Structure

Each project is organized in its own folder with:

• README.md – Project overview & key insights
• Jupyter Notebook / Python Scripts – Code with step-by-step analysis
• Data – Link to or sample datasets used
• Visuals / Output – Plots, model evaluations, dashboards (if any)

---

🌟 Featured Projects

Project	Machine Learning Technique Used	Description
Breast Cancer SVM Classification	Support Vector Machine (SVM)	Predict breast cancer diagnosis using SVM.
COVID-19 Dashboard	Data Visualization, API Integration	Interactive global tracker for COVID-19 trends using Plotly Dash and real-time data.
Credit Card Fraud Detection	Random Forest, XGBoost	Detect fraudulent transactions using Random Forest and XGBoost with an emphasis on class imbalance.
Customer Segmentation	K-Means Clustering	Segment customers based on purchasing behavior using unsupervised learning (K-Means).
Customer Churn Prediction	Logistic Regression, Random Forest	Predict customer churn using classification algorithms like Random Forest and Logistic Regression.
Diabetes Prediction with Naive Bayes	Naive Bayes Classifier	Predict the likelihood of diabetes using Naive Bayes with health metrics.
Diabetes Prediction with Neural Networks	Neural Networks (Deep Learning)	Predict diabetes using deep learning models and compare with traditional methods.
Gene Expression Analysis with PCA, t-SNE, UMAP	PCA, t-SNE, UMAP	Visualize gene expression data using dimensionality reduction techniques (PCA, t-SNE, UMAP).
House Price Prediction	Linear Regression, XGBoost	Predict house prices using regression models and advanced techniques like XGBoost.
Hugging Face Bioinformatics	NLP, Transformers (Hugging Face)	Apply NLP models (Hugging Face transformers) for bioinformatics tasks like protein sequence analysis.
Machine Learning Model Evaluation	Model Evaluation Metrics (e.g., ROC-AUC, Confusion Matrix)	Visualize and compare model evaluation metrics for classification models.
Metagenomics Taxonomy Classification	Multi-Class Classification, Random Forest	Classify taxonomic groups from metagenomic sequence data using Random Forest.
ML Model from Scratch	Linear Regression, Logistic Regression	Build machine learning algorithms like linear and logistic regression from scratch using Python.
ML Project Template	Project Organization	Reusable template for organizing machine learning projects in a structured and reproducible way.
NLP Spam Classifier	Naive Bayes, TF-IDF	Classify spam emails using Naive Bayes and TF-IDF for text classification.
Parkinson's Disease Prediction	Binary Classification, Random Forest	Predict Parkinson’s disease using voice data analysis and classification algorithms like Random Forest.
Resume Screening using GPT & LangChain	NLP, GPT, LangChain	Automate resume screening using GPT and LangChain integration for information extraction.
Sepsis Detection from ICU Data	Time-Series Analysis, XGBoost	Detect early signs of sepsis using ICU patient time-series data with XGBoost.
SQL-based ML Feature Engineering	SQL, Feature Engineering	Generate ML features directly from SQL queries for structured databases.
Statistical Analysis Notebook	Statistical Inference, Hypothesis Testing	Explore core statistical concepts like hypothesis testing and distributions using Python libraries.
Survival Analysis	Kaplan-Meier, Cox Regression	Analyze time-to-event data using survival analysis techniques like Kaplan-Meier and Cox regression.
Transfer Learning with ResNet	Transfer Learning, CNNs	Use ResNet for transfer learning on image classification tasks with pre-trained models.
XGBoost Binary Classification	XGBoost, Binary Classification	Perform binary classification using XGBoost with a focus on hyperparameter tuning.
XGBoost Regression	XGBoost Regression	Predict continuous values using XGBoost regression models with feature selection and regularization.
MXNet CNN Image Classification	Convolutional Neural Networks (CNNs), MXNet	Use MXNet to build CNN-based models for image classification tasks.
Real-Time Image Classification CNN	Convolutional Neural Networks (CNNs)	Build a real-time image classification system using CNNs for visual recognition tasks.
COVID-19 Bayesian Forecasting	Bayesian Statistics, Time-Series Forecasting	Use Bayesian techniques for forecasting COVID-19 trends and uncertainties.
Gene Expression Forecasting ARIMA SARIMA	ARIMA, SARIMA	Forecast gene expression levels using time-series models like ARIMA and SARIMA.
Protein Secondary Structure Prediction (Transformer)	Transformer Networks	Predict protein secondary structures using Transformer-based deep learning models.
Titanic Decision Tree	Decision Tree	Predict passenger survival on the Titanic using a decision tree classifier.
Titanic Fare Prediction	Linear Regression, Random Forest	Predict the fare price of Titanic passengers based on various features using regression models.
Titanic Survival Prediction Logistic Regression	Logistic Regression	Predict Titanic passenger survival using logistic regression based on personal and travel information.
Customer Segmentation K-Means	K-Means Clustering	Segment customers using unsupervised learning with K-Means clustering based on purchasing behaviors.
LightGBM Text Classification	LightGBM, Text Classification	Build a text classification model using LightGBM for fast and efficient training on large datasets.
Random Forest ML Classification	Random Forest, Classification	Classify data into categories using Random Forest with various feature engineering techniques.
Random Forest Regression	Random Forest, Regression	Predict continuous outputs using Random Forest regression models.
Deep Learning Protein Structure Prediction	Deep Learning, CNN, LSTM	Use deep learning models like CNN and LSTM for predicting protein structures from sequences.
MLflow XGBoost	XGBoost, MLflow	Build and track XGBoost models with MLflow for scalable machine learning experiments.

---

🔍 Machine Learning Techniques

Supervised Learning

• Logistic Regression, Random Forest, Decision Tree, XGBoost, Ridge, Lasso

Unsupervised Learning

• K-Means, DBSCAN, PCA

Natural Language Processing (NLP)

• Tokenization, TF-IDF, Naive Bayes

Model Evaluation

• Confusion Matrix, ROC-AUC, F1-Score, Cross-validation

Hyperparameter Tuning

• GridSearchCV, RandomizedSearchCV

---

🛠️ Tools & Technologies

Languages

• Python, SQL

Libraries

• pandas, numpy, scikit-learn, matplotlib, seaborn, XGBoost, NLTK

Visualization

• Plotly, Dash, Tableau

Pipelines

• Jupyter, Airflow, WDL, Azure Data Factory

Others

• Git, VSCode, GitHub Actions

---

👨‍💻 About Me

I’m Manish Kumar, a data professional with 16+ years of experience in bioinformatics software development and data engineering. I'm now applying that experience to broader data science and machine learning domains, with a passion for solving real-world problems using data-driven solutions.

---

📬 Contact

• Email: mandecent.gupta@gmail.com
• LinkedIn: linkedin.com/in/manish-kumar-0160837

If you like this repo, please consider ⭐ starring it and connecting with me on LinkedIn!