Credit Scoring and Data Science - Job Description
Function: Analytics
Positions: Data Scientists (2)
Job Purpose
Data scientists are responsible for developing machine learning and advanced statistical models on banking datasets for the credit risk and marketing functions
Broad Areas
Key Responsibilities
- Model Development & Knowledge Management - Predictive Analytics and Machine learning
- Developing decision support systems for banks using predictive analytics techniques and machine learning models
- Designing features and predictor variables through feature engineering
- Selecting features, building and optimizing classifiers using advanced analytics and machine learning techniques
- Using R / Python for data cleansing and preparation for analysis and models development
- Maintaining knowledge repository of models to automate repetitive tasks and functions
- Enhancing data collection procedures within the banks to include information that is relevant for building analytic systems
- Extending the client bank's data with third party data sources and alternate data sources to improve predictive power of the models
- Developing credit scoring and propensity models using traditional scoring methodologies
- Enhancing the model performance through machine learning and deep learning methods
- Credit Scoring and Modelling Process Automation, and Model Enhancement Through Alternate Data Sources
- Designing codes for data cleansing on R / Python to automate data preparation and pre-processing activities
- Automating repetitive scoring tasks for fast delivery
- Working on integrating the alternate and unstructured data sources (text, images, narratives etc.) to enhance the model performance
KNOWLEDGE AND SKILLS (Knowledge and skills needed for satisfactory performance of the job)
- Postgraduate in quantitative disciplines i.e. statistics, machine learning, physics, operation research, applied mathematics, engineering etc.
- Through working knowledge of statistical/machine learning techniques like logistic regression, support vector machines, random forests, gradient boosting, na- ve Bayes, neural networks, deep learning, recurrent neural network (RNN), natural language processing (NLP), k-mean clusters, etc.
- Minimum 1 year of experience of working in banks / financial institutions
- Good working knowledge of credit scoring and development methodology
- 3-4 years of cumulative experience on R, Python SciKit-learn, or SAS
- Good experience of feature engineering and feature selection
- At least 2 years of experience on R or Python SciKit-learn / Pandas / Tensor Flow
- Good communication skills
Didn’t find the job appropriate? Report this Job