Visibility distance prediction

Problem Statement :-

The objective of this project is to develop a machine learning model that can accurately predict the maximum visibility distance in a given location and weather condition. The model should take into account various weather parameters such as humidity, temperature, wind speed, and atmospheric pressure, as well as geographical features such as elevation, terrain, and land cover. The model should be trained on a large dataset of historical weather and visibility data, and validated using a separate test dataset. The ultimate goal is to provide a tool that can help improve safety and efficiency in various applications such as aviation, transportation, and outdoor activities.

Tech Stack Used

1. Python
2. FastAPI
3. Machine learning algorithms
4. Docker
5. MongoDB

Infrastructure required

1. AWS S3
2. Azure
3. Github Actions

How to run

Before you run this project make sure you have MongoDB Atlas account and you have the shipping dataset into it.

Step 1. Cloning the repository.

git clone https://github.com/Machine-Learning-01/Customer_segmentation.git

Step 2. Create a conda environment.

conda create --prefix venv python=3.7 -y

conda activate venv/

Step 3. Install the requirements

pip install -r requirements.txt

Step 4. Export the environment variable

export AWS_ACCESS_KEY_ID=<AWS_ACCESS_KEY_ID>


export AWS_SECRET_ACCESS_KEY=<AWS_SECRET_ACCESS_KEY>


export AWS_DEFAULT_REGION=<AWS_DEFAULT_REGION>


export MONGODB_URL= <MONGODB_URL>

Step 5. Run the application server

python app.py

Step 6. Train application

http://localhost:5000/train

Step 7. Prediction application

http://localhost:5000/predict

Run locally

1. Check if the Dockerfile is available in the project directory
2. Build the Docker image

docker build --build-arg AWS_ACCESS_KEY_ID=<AWS_ACCESS_KEY_ID> --build-arg AWS_SECRET_ACCESS_KEY=<AWS_SECRET_ACCESS_KEY> --build-arg AWS_DEFAULT_REGION=<AWS_DEFAULT_REGION> --build-arg MONGODB_URL=<MONGODB_URL> . 

3. Run the Docker image

docker run -d -p 5000:5000 <IMAGE_NAME>

Project Architecture -

Data Collection Architecture -

Deployment Architecture -

Models Used

• K-Means
• LogisticRegression

From these above models after hyperparameter optimization we selected these two models which were K-Means for clustering and Logistic Regression for classification and used the following in Pipeline.

• GridSearchCV is used for Hyperparameter Optimization in the pipeline.

src is the main package folder which contains

Components : Contains all components of Machine Learning Project

• Data Ingestion
• Data Validation
• Data Transformation
• Data Clustering
• Model Trainer
• Model Evaluation
• Model Pusher

Custom Logger and Exceptions are used in the Project for better debugging purposes.

Conclusion

• This Project can be used in real-life by Users.