Karaj Khosla

Developed algorithms to optimize black-box systems using user feedback from pairwise comparisons. Employed preference learning, black-box optimization, and surrogate modeling techniques (RBF, Gaussian processes) to efficiently explore the parameter space and identify optimal configurations.

- At the Institut für Energiewirtschaft und Rationelle Energieanwendung (IER). - Led data science lifecycle for complex energy data, including data cleaning, EDA, feature engineering, modeling, and deployment. Developed software for efficient energy data management and analysis. Applied machine learning to extract insights and created visualizations for data-driven energy decision-making. Communicated findings effectively to technical and non-technical audiences.

- At the Visualisierungsinstitut der Universität Stuttgart (VISUS). - Implemented Python package for Uncertainty-Aware-MDS and PCA, novel dimensionality reduction algorithms. Ensured code quality, efficiency, and testing. Stayed updated on dimensionality reduction and uncertainty quantification research. Maintained detailed project documentation.

-Implemented Prediction using Supervised ML, Unsupervised ML, and Decision Tree Algorithm. -Performed Exploratory Data Analysis on a variety of Datasets including Retail, Terrorism, and Sports.

- Developed Python-based Software to Automate processes like Email Marketing and a variety of Social Media marketing tools. -Designed and executed automation within software processes.

- Developed a method better than tSNE, which is an essential tool for Dimension Reduction and Data Visualization in Machine Learning across the World. - Worked on a Novel Algorithm for Data Analysis and Data Visualization for Big Data using Statistics and Machine Learning. - Research Paper is Published in MDPI Algorithms.