Build Propensity Model to Predict POS Adoption by SME Clients (Python / ML)
Budget: $250 – $750 USD
Seeking a skilled data scientist or machine learning expert to build a predictive model that identifies SMEs (small and medium enterprises) with a high propensity to adopt a POS (Point-of-Sale) system in the next 3 months.
This project simulates a real banking scenario where CRM and commercial teams must prioritize leads. The goal is to support the bank’s salesforce by ranking SME clients using a model that predicts whether they are likely to contract a POS terminal based on a rich dataset of company behaviors and financial indicators.
**Deliverables:
1.Data Cleaning & Exploration
Analyze training dataset to understand patterns
Perform preprocessing (e.g., missing values, outliers, feature engineering)
2.Model Development
Supervised binary classification model (e.g., Logistic Regression, Random Forest, XGBoost)
Evaluate using Accuracy, Precision, Recall, and F1 Score
Explanation of model selection and feature importance
3.Prediction File
A .txt file with two columns: id;prediction using the test dataset
Format must match: 123456;1 where 1 indicates POS adoption
4.Report for Response Template
Short written answers for five questions (provided in Word doc)
Explain key variables, model logic, data preparation, and improvement ideas
***Files Provided:
Training & test datasets
Data dictionary
Assignment instructions
Report template to be filled in
**Skills Required:
Python (Pandas, Scikit-learn, or equivalent)
Data preprocessing & feature selection
Binary classification modeling
Model evaluation & business interpretation
***Timeline:
Expected delivery within 10–12 days
This project simulates a real banking scenario where CRM and commercial teams must prioritize leads. The goal is to support the bank’s salesforce by ranking SME clients using a model that predicts whether they are likely to contract a POS terminal based on a rich dataset of company behaviors and financial indicators.
**Deliverables:
1.Data Cleaning & Exploration
Analyze training dataset to understand patterns
Perform preprocessing (e.g., missing values, outliers, feature engineering)
2.Model Development
Supervised binary classification model (e.g., Logistic Regression, Random Forest, XGBoost)
Evaluate using Accuracy, Precision, Recall, and F1 Score
Explanation of model selection and feature importance
3.Prediction File
A .txt file with two columns: id;prediction using the test dataset
Format must match: 123456;1 where 1 indicates POS adoption
4.Report for Response Template
Short written answers for five questions (provided in Word doc)
Explain key variables, model logic, data preparation, and improvement ideas
***Files Provided:
Training & test datasets
Data dictionary
Assignment instructions
Report template to be filled in
**Skills Required:
Python (Pandas, Scikit-learn, or equivalent)
Data preprocessing & feature selection
Binary classification modeling
Model evaluation & business interpretation
***Timeline:
Expected delivery within 10–12 days