- Developed a classifier to identify and predict fraudulent financial activities combining a foundatioanl KMeans
- Achieved a 25% increase in F1 score(from 0.68 to 0.85) in the anomaly detection suite and a 15% reduction
- 10% from 12%) in false positive rate over baseline models
- Tools and Tech.: Python, PyTorch, AWS SageMaker, MLFlow
K
Kroop AI
Deep Learning Intern
Gandhinagar, IN-GJ, IN
August 2021 - December 2021
I
Indian Institute of Management
Research Intern
Ahmedabad, IN-GJ-AB, IN
May 2021 - August 2021
Skills
3d ModelsAlgorithmsAmazon Elastic Compute CloudAmazon Web ServicesAnalytical ThinkingAnomaly DetectionApache SparkArtificial IntelligenceCluster AnalysisCMakeC++ (Programming Language)C Sharp (Programming Language)DatabasesDeep LearningDockerFundraisingGame TheoryGenerative AIGitInformation TechnologyKerasKnowledge of Electrical EngineeringKnowledge of FinanceKnowledge of MathematicsKnowledge of StatisticsLeadershipLinear AlgebraMachine LearningMySQLNumerical AnalysisNvidia CUDAOpenCVPredictive ModellingPython (Programming Language)PytorchRecommender SystemsRisk AnalysisSchedulingSelf MotivationSemanticsSocial WorkSoftware EngineeringSQL DatabasesStochastic ProcessTeachingTensorflow