SquadStack aims to help businesses improve their sales and customer engagement by providing a blend of human expertise and AI-driven technology. Their platform optimizes sales processes through efficient lead management, intelligent outreach, and data-driven insights to maximize conversions and productivity.
With keeping this aim in mind this project is a whole workflow that starts from data ingestion all the way upto predicting unseen values with our model
The purpose of this repository is to demonstrate the data analysis and the model development that took part to complete this project
Let's jump into the Python packages you need. Within the Python environment of your choice, run:
git clone https://github.com/hs414171/Squadstack_Activity
pip install -r requirements.txt
To initiate our project, we will develop a model using the classification Machine Learning algorithms designed to predict wether the call was converted or not. This model uses the following Algorithms: -
- Logistic regression
- Random Forest Classifier
- XGBoost Classifier
Our standard training pipeline consists of several steps:
- ingest_data: This step will ingest the data and create a DataFrame.
- handle_values : This handles all the missing values and removes columns that are not required
- split_data: This step will split the data into train test and validation sets
- handle_imbalance : This handles the class imbalance using various techniques like SMOTE or Random Under Sampling
- train_model: This step will train the model and save the model.
- evaluation: This step will evaluate the model and save the metrics.
This could be termed as the training pipeline which runs all the pipeine steps in a defined workflow and saves the results such as the trained models, predictions made etc.
You can run workflow as follows:
run_workflow.py
python run_workflow.py
@hs414171