This project aims to create a web interface that allows users to predict doctor consultation fees based on their input. The machine learning model was trained on a dataset obtained by scraping data from the Practo website using Selenium. With the use of Python Pandas, the scraped data was thoroughly cleaned & preprocessed for accurate predictions.
| Files/Folder | Description |
|---|---|
| Dataset Folder | This folder provides data state wise in csv format |
| Python File | This contains the .ipynb file of the analysis for Data Extract, Data cleaning, EDA and ML Models. |
| HTML File | This contains the .html file for User Interface. |
| PowerPointPresentation | This file provides the PowerPoint presentation which contains all the major insights and conclusions. |
- According to the Practo dataset, Bangalore has the highest number of doctors.
- The most common degrees among doctors are MBBS, MD, and BDS, with the highest representation in the dataset.
- The dataset indicates that the three most prominent specialties among doctors are:
- Dentist
- Gynecologist
- Pediatrician
1. Data Collection: Gathered doctor-related information from Practo using web scraping techniques with Selenium.
2. Data Preprocessing: Conducted thorough data cleaning, handling missing values, and transforming categorical variables into numerical representations.
3. Feature Engineering: Derived additional relevant features from the existing dataset, such as extracting qualifications.
4. Model Selection: Explored various regression algorithms and selected potential candidates based on initial performance evaluation.
5. Hyperparameter Tuning: Utilized GridSearchCV to fine-tune hyperparameters for each selected model, optimizing their performance.
6. Model Training: Trained multiple models on the dataset with the tuned hyperparameters to improve predictive accuracy.
7. Weighted Voting: Implemented a Weighted Voting technique, combining predictions from multiple models, each with a specific weight.
8. Model Evaluation: Evaluated the ensemble model using appropriate metrics such as Mean Absolute Error (MAE) or Root Mean Squared Error (RMSE) to measure prediction accuracy.
9. Web App Development: Developed a user-friendly web application using Flask, HTML, and CSS to offer an intuitive interface for users to input parameters.
1. Feature Engineering: Handled complex features, especially those with diverse and numerous categories.
2. Model Selection: Explored different ML models to identify the Best models.
3. Hyperparameter Tunning: Hyperparameter tuning was time-consuming due to limited time for model development
4. Model Deployment: Explored model deployment options.
1. Healthcare Accessibility: By giving patients an idea of potential costs, it helps them seek appropriate medical care without the barrier of uncertainty about fees.
2. Transparency and Trust: Clear fee estimates foster trust and confidence in medical services, enhancing the doctor-patient relationship.
3. Efficiency for Providers: With fee estimates readily available, administrative processes become smoother, leading to improved overall service efficiency.