Gangadhar Kurma

1. Extracting data from the database.
2. On the extracted data, I did data preprocessing and the feature engineering.
3. Train the data with different machine learning algorithms and choose the best one based on metrics.
4. Save the model.
5. Exploring new technologies comes in my domain.
6. Try to work on different research areas.

Implemented Variational Autoencoders (VAEs) using the Keras framework for one-class classification, focusing on healthcare and car insurance datasets.
Utilized Chaos numbers instead of normal distribution during training, enhancing the model’s performance and robustness.
Successfully applied VAEs to the datasets, achieving improved anomaly detection and classification accuracy.
Demonstrated strong technical skills in deep learning, unsupervised learning, VAE architecture, and training processes.
Showcased the ability to analyze complex datasets, extract meaningful insights, and make data-driven decisions for real-world applications.
Set up an on-premise Yarn cluster for the Hadoop/Spark ecosystem, enabling efficient Big Data analytics and processing.
Collected data from various sources and performed data preprocessing using PySpark, ensuring data quality and consistency.
Applied unsupervised machine learning algorithms like FP-Growth from the MLlib package to uncover association rules and patterns in the data.
Developed distributed computing code for the Apriori algorithm, facilitating scalable and efficient association rule mining.
Analyzed and evaluated model performance using key metrics such as support, confidence, and lift, providing valuable insights for data-driven decision-making in association rule mining.
Employed LIME and SHAP techniques to enhance the interpretability of ML models for banking churn prediction and NPA loan default data.
Explained the decision-making process of complex AI models such as Logistic Regression and Decision Trees using XAI methods.
Established transparency and accountability in the use of AI systems by making black-box decisions more comprehensible.
Minimized the impact of model bias and reduced the cost of mistakes by using XAI techniques.
Analyzed and evaluated the performance of ML models using XAI methods and provided recommendations for improvement.
Developed and implemented XAI solutions to address model errors and improve decision-making in the banking industry.
Collaborated with stakeholders to ensure compliance with ethical and legal standards in the use of AI systems.
Presented XAI findings and recommendations to decision-makers in the banking industry to inform strategic decision-making.
Developed a sentiment analysis model using PySpark to analyze and classify customer reviews of a banking app as positive, negative, or neutral.
Cleaned and pre-processed the raw data by removing stop words, punctuations, and special characters, and performed tokenization and stemming.
Generated word clouds and plotted frequency distribution graphs to visualize the most common positive and negative keywords used in the reviews.
Built a sentiment analysis model using machine learning algorithms like logistic regression and random forest, and evaluated the model’s accuracy using metrics like precision, recall, and F1-score.
Deployed the sentiment analysis model within the premises of the organization.
Improved the overall customer experience of the banking app by identifying the most common pain points and addressing them proactively.
Proficient in big data technologies and experienced in setting up and maintaining on-premises Hadoop, Spark, and Yarn clusters to process and analyze large-scale data.
Developed and deployed a heart disease prediction model using machine learning algorithms like Random Forest, Decision Tree, KNN, and Logistic Regression.
Built a web application using Flask, HTML, and CSS for real-time predictions.
Proficient in Python, Flask, HTML, and CSS, learning algorithms, data pre-processing, feature engineering, model selection and evaluation, and web application development.
Cleaned and pre-processed the raw medical data, handled missing values, performed feature scaling, and engineered new features to improve model performance.
Trained and evaluated multiple machine learning models using techniques like cross-validation, hyperparameter tuning, and feature selection, and selected the best-performing model based on metrics like accuracy, precision, recall, and F1-score.
Built a user-friendly web application using Flask, HTML, and CSS, and integrated the heart disease prediction model to enable real-time predictions based on user input.
Deployed the heart disease prediction model and the web application locally using Docker and Kubernetes, as well as in cloud platforms like AWS, ensuring scalability, performance, and security of the application.

Successfully developed a wine quality prediction system using machine learning techniques.
Utilized the pandas profiling report to perform in-depth analysis of the dataset, gaining valuable insights into its characteristics.
Explored and evaluated multiple machine learning models, ultimately achieving high accuracy scores with the XGBoost classifier.
Selected the XGBoost classifier as the final model due to its superior predictive power and accuracy in classifying wine quality.
Saved the trained XGBoost classifier model for future use, enabling efficient and accurate predictions on new, unseen wine samples.
Developed an Optical Character Recognition (OCR) system to extract text from images, focusing on accurately printing the text present in the images.
Utilized the PyTesseract package, a powerful OCR library, to achieve robust and accurate text recognition results.
Implemented preprocessing techniques to enhance image quality and improve OCR accuracy, ensuring reliable text extraction.
Developed a predictive model to simultaneously predict employee salary and satisfaction levels.
Divided the dataset into two subsets: one for regression to predict salary as a continuous variable and another for classification to predict employee satisfaction.
Employed regression models for the salary dataset and classification models for the satisfaction dataset.
Identified Support Vector Machines (SVM) as the best-performing model for both regression and classification tasks.
Demonstrated expertise in predictive modeling, regression analysis, classification algorithms, and the effective use of SVM for accurate salary and satisfaction predictions.