Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/shibbir-ahmad24/water-potability-prediction-using-hybrid-machine-learning-techniques
Water Potability Prediction Using Hybrid Machine Learning Techniques
https://github.com/shibbir-ahmad24/water-potability-prediction-using-hybrid-machine-learning-techniques
machine-learning matplotlib numpy pandas python3 scikit-learn seaborn
Last synced: 29 days ago
JSON representation
Water Potability Prediction Using Hybrid Machine Learning Techniques
- Host: GitHub
- URL: https://github.com/shibbir-ahmad24/water-potability-prediction-using-hybrid-machine-learning-techniques
- Owner: shibbir-ahmad24
- Created: 2024-06-11T00:31:40.000Z (7 months ago)
- Default Branch: main
- Last Pushed: 2024-08-31T22:46:27.000Z (5 months ago)
- Last Synced: 2024-10-26T22:52:27.953Z (3 months ago)
- Topics: machine-learning, matplotlib, numpy, pandas, python3, scikit-learn, seaborn
- Language: Jupyter Notebook
- Homepage:
- Size: 6.01 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Water Potability Prediction Using Hybrid Machine Learning Techniques
# **Problem**
Access to safe drinking water is a fundamental human right, yet millions of people worldwide lack reliable access to potable water. Contaminated water poses significant health risks, leading to diseases and even death. Ensuring water safety requires robust predictive models to determine potability based on various chemical and physical properties of water.
# **Solution**
The project aimed to develop a hybrid machine learning model that accurately predicts water potability. Multiple algorithms were employed, including Logistic Regression, Random Forest, LightGBM, K-Nearest Neighbors (KNN), and a custom Hybrid Model (Random Forest + LightGBM). The models were trained using historical water quality data and validated on test dataset to assess their performance.
# **Impact**
The developed Hybrid Model achieved the highest accuracy, precision, recall, and F1-score, significantly improving water potability predictions. This model can be used by water safety authorities to assess the safety of drinking water and take preventive measures, thereby reducing the risk of waterborne diseases and enhancing public health.
# **Technology Used**
- Python
- Machine Learning
- Numpy
- Pandas
- Scikit-learn
# **Dataset**
- [Dataset](https://www.kaggle.com/datasets/adityakadiwal/water-potability/data)
- **Features Description:**
1. **pH**: pH of water (0 to 14).
2. **Hardness**: Capacity of water to precipitate soap in mg/L.
3. **Solids**: Total dissolved solids in ppm.
4. **Chloramines**: Amount of Chloramines in ppm.
5. **Sulfate**: Amount of Sulfates dissolved in mg/L.
6. **Conductivity**: Electrical conductivity of water in μS/cm.
7. **Organic_carbon**: Amount of organic carbon in ppm.
8. **Trihalomethanes**: Amount of Trihalomethanes in μg/L.
9. **Turbidity**: Measure of light emitting property of water in NTU.
10. **Potability**: Indicates if water is safe for human consumption. Potable - 1 and Not potable - 0.
# **Method**
- **Data Load**: Imported the dataset into Python using Pandas.
- **EDA (Exploratory Data Analysis)**: Analyzed the dataset for missing values, and feature distributions.
- **Data Cleaning**: Handled missing values, ensuring data quality.
- **Data Sampling**: Applied random undersampling to balance the classes in the dataset.
- **Normalization**: Scaled the feature data to improve model performance.
- **Data Splitting**: Split the dataset into training and testing sets.
- **Model Training**: Trained multiple machine learning models on the training dataset.
- **Model Testing**: Evaluated the models on the testing dataset using various performance metrics.
- **Comparative Result Analysis**: Compared the models based on accuracy, precision, recall, and F1-score.
- **Summary**: Highlighted the best-performing model and discussed its potential applications.
# **Analysis Visuals**
# **Comparative Results Analysis**
### **Accuracy:**
The **Random Forest**, **LightGBM**, and **Hybrid Model** exhibit the highest accuracy (0.62), outperforming **Logistic Regression** (0.48) and **KNN** (0.51).
### **Precision**
The **Hybrid Model** leads in precision (0.62), slightly surpassing **Random Forest** (0.58) and **LightGBM** (0.59). **Logistic Regression** and **KNN** trail behind with 0.46 and 0.48, respectively.
### **Recall**
The **Hybrid Model** achieves the highest recall (0.62), matching **Random Forest** and slightly exceeding **LightGBM** (0.60). **Logistic Regression** shows a higher recall (0.59) compared to **KNN** (0.47).
### **F1-Score**
The **Hybrid Model** has the highest F1-Score (0.62), followed by **Random Forest** and **LightGBM** (both 0.60). **Logistic Regression** and **KNN** score lower at 0.51 and 0.47, respectively.
# **Summary**
The **Hybrid Model** outperforms other models across all performance metrics, closely followed by **Random Forest** and **LightGBM**. **Logistic Regression** and **KNN** underperform in comparison, particularly in terms of accuracy, precision and F1-Score. The **Hybrid Model**'s robust performance suggests its potential for real-world applications in water quality assessment and safety assurance.