Timeline: September 2024 – December 2024
Tech Stack: Streamlit · Python · Scikit-learn · TensorFlow/Keras · Pandas · NumPy · MongoDB
An end-to-end ML & DL workflow automation platform built with Streamlit that empowers users to train, evaluate, visualize, and deploy machine learning models seamlessly — without writing code. It supports classification, regression, and clustering tasks, making model experimentation and insight generation accessible for both technical and non-technical users.
- 🔧 Interactive Workflow UI via Streamlit
- 🤖 Supports Classical ML (Scikit-learn) and Deep Learning (TensorFlow/Keras) models
- 📊 Real-time Data Visualization and Metric Reports
- 🧠 Train models for:
- Classification (Logistic Regression, Random Forest, DNN)
- Regression (Linear Regression, SVR, DNN)
- Clustering (K-Means, DBSCAN)
- 📁 Upload your dataset (CSV format)
- 🔍 Data Preprocessing pipeline (null handling, encoding, scaling)
- 📈 Model Evaluation (confusion matrix, accuracy, MSE, silhouette score, etc.)
- 💾 MongoDB for storing model performance metadata & logs
- 🚀 Model deployment-ready interface with experiment tracking
git clone https://github.com/your-username/automl-streamlit-app.git
cd automl-streamlit-app- Install Dependencies
pip install -r requirements.txt- Launch the App
streamlit run app.py
🛠️ ML/DL Models Used Task Algorithms Used Classification Logistic Regression, Random Forest, SVM, KNN, Deep Neural Network (Keras) Regression Linear Regression, Decision Tree Regressor, SVR, DNN Regressor (Keras) Clustering KMeans, DBSCAN, Hierarchical Clustering
💡 Sample Use Case Flow Upload Dataset
Data Preview & Clean
Choose ML/DL Task: Classification / Regression / Clustering
Train Model (customize hyperparameters)
Evaluate & Visualize Results
Save and Log Model Info in MongoDB
📦 MongoDB Integration Stores experiment metadata:
Model type
Accuracy / Loss
Hyperparameters
Timestamp
Easily retrieve past experiment logs for reproducibility and tracking