Introduction to Make-a-thon: In the competition, participants are challenged to develop a lead-scoring algorithm that can accurately predict the likelihood of a customer making a purchase on an e-commerce platform based on accuracy and recall value.
Recall value is an essential metric in lead scoring, as it measures the algorithm's ability to correctly identify all positive cases (i.e., customers who are likely to make a purchase) out of all actual positive cases in the dataset. In other words, it measures how well the algorithm is able to capture all potential customers who are likely to make a purchase.
The competition judges the participants' algorithms based on both accuracy and recall value. The winning entry is the one that provides the most accurate predictions while also achieving the highest recall value. This ensures that the algorithm is accurate and effective at identifying potential customers who are likely to make a purchase. The competition is designed to provide valuable insights and solutions to e-commerce businesses looking to increase their sales and revenue by accurately identifying potential customers who are likely to make a purchase.
Tools & Technologies to be Used in Competition: Python, Jupyter Notebook
Reference & Inspiration for Participants: There are various prebuilt models based on lead-scoring algorithms. When developing a lead-scoring model, it can be helpful to keep in mind prebuilt lead-scoring companies that offer similar services. These companies can provide inspiration for your algorithm development and help you to understand the types of features and techniques that are commonly used in the industry. Some examples of prebuilt lead-scoring companies include:
Infer: Infer offers a lead-scoring platform that uses predictive analytics and machine learning to score leads based on their likelihood to convert.
MadKudu: Their platform provides a simple dashboard that allows sales and marketing teams to easily prioritize their leads based on their likelihood to convert.