PRASHANTHI GUNJAPALLI

Highly-motivated employee with desire to take on new challenges. Strong worth ethic, adaptability and exceptional interpersonal skills. Adapt at working effectively unsupervised and quickly mastering new skills.

• Tools: Xilinx Vivado, Questa, Xcelium, VCS, Riviera, Jupyter Notebook, Linux.
• Programming Languages: PERL, TCL, XML, PYTHON.
• Packages: Numpy, Pandas, Sklearn, Scipy, MatplotLib, StatsModel.
• Satistics/MachineLearning: Statistical Analysis, Linear/Logistic Regression, Clustering, Regularization.

• Trained in Static Timing Analysis [June'18 - Nov'18]
• Licensed Silicon Design Engineer, AMD-Xilinx [Nov'18 - Current]
• Certified Post Graduate in Artificial Intelligence and Machine Learning, The University of Texas, Austin & GREAT LAKES [June'2022 - June'2023]
• Microsoft Certified: Azure AI Fundamentals, 2023

MACHINE LEARNING PROJECTS:

• Applied Statistics - Based on the report on the performance shown by the BasketBall teams, predicting the teams which will be a deal win for the upcoming tournament using probabilistic approaches by performing detailed statistical analysis and EDA using uni-variate, bi-variate and multi-variate techniques.
• Hypothesis Testing - Analyzed the attributes given by various companies to draw conclusions on the funds raised by the companies which are still operating vs. the companies which are closed down.
• Supervised Learning - Trained a KNN Classifier model to be able to generate predictions on the patients with certain conditions given the biomechanics features derived from the shape and orientation of the body part. Built a Logistic Regression model to perform focused marketing by predicting the potential customers where in the bank can increase the conversion ratio to double digit with same budget. Also trained the model in SVM for the same and analyzed the evaluation metrics for both.
• Ensemble Techniques - Aimed and XGBoost Classifier model that will help to identify the potential customers who have a higher probability to churn which helps the company to understand the pinpoints and patterns of customer churn and will increase the focus on strategizing customer retention.
• Unsupervised Learning - Understanding the K-Means Clustering by applying on the given Car Dataset with multivalued discrete and continuous attributes to segment the cars into various attributes. Classified a given silhouette as one of three types of vehicle, using a set of features extracted from the silhouette by applying dimensionality reduction technique - PCA and trained the model with SVM.
• Featurization, Model Selection & Tuning - By applying Cross validation techniques such as LOOCV, K-Fold, GridSearch CV, Standardization/Normalization on the given dataset in which a semiconductor manufacturing company require a Pass/Fail yield of a particular process entity. Best accuracy was given by the KNN classifier.
• Recommendation Systems - Built a recommendation system using popularity based and collaborative filtering methods to recommend mobile phones to a user which are most popular and personalized respectively.

DEEP LEARNING PROJECTS:

• Introduction to Neural Networks & Deep Learning - The project was accomplished by delivering 2 sub-projects. Part 1: Deploys a neural network to build a regressor & classifier respectively for a communications equipment manufacturer. The model predicts the equipment’s signal quality using various parameters from its products, which is responsible for emitting informative signals. Part 2: Delivers an image classifier, which can classify numbers from the photographs captured at street level using a Neural Network .
• CNN Architecture and Transfer Learning (Computer Vision) - This project involves 2 subprojects to solve the problem of a botanical research group. Part 1: Image classifier capable of determining a plant's species and detailed analysis on how CNN is a better image classifier over traditional methods. Part 2: Image classifier capable of determining a flower’s species using CNN and curating an image dataset. Dataset - The dataset comprises images from 12 plant species. The dataset comprises images from 17 flower species “import tflearn.datasets.oxflower1" .
• Object Detection and Recognition (Computer Vision) - The project involves 3 sub segments Part 1 Implement an object detection model for highlighting human faces to automate the process of providing information of cast and crew while streaming. Part 2 Curate a training dataset to be used for highlighting human faces Part 3 Implement a face identification model for a company, which intends to recognize human faces from images.
• Sequential Natural Language Processing - This project involves delivering 2 Sub projects based on sequential NLP. Part 1: Implementing a sequential NLP based text classifier, for a company in digital content & entertainment industry, using input text parameters to determine the customer's sentiments based on customer reviews in IMDB database Part 2: Implementing a sequential NLP based text classifier for a social media analytics company using input text parameters to determine the customer sentiments for sarcasm detection.

CAPSTONE PROJECT:

       The capstone project is a focused approach to attempt a real-life challenge with the learnings from the post graduation program done with Great Learning. 

 With a group of four people, we tried to develop a model using Natural Language procession techniques which can output the severity level of Accidents arise Industries.Companies given the description of the accident. 

A brief summary of the project:  With the given input features and the target variable of an Industrial Data, we have trained a variety of Machine Learning (such as Logistic Regression, KNN, Random Forest, SVC, Decision Tree) and Deep Learning algorithms (ANN, CNN, RNN, LSTM). However, LSTM has performed well on both the training and test data to predict the severity of accident levels for a given description.