Krishna Sumanth Mannala

Have 3 years and 7 months of experience. Worked in diverse fields such as oracle databases, data analysis, and machine learning. Currently pursuing master's degree.

• Automated Metric Collection and Analysis:-We need to periodically collect CPU, memory, IO, and storage utilization details of approximately 24 databases/servers and prepare a report. I used requests to automatically download the CPU, memory utilization csv files and cx_oracle to connect to invidual databases and fetch IO, storage data. I used matplotlib and seaborn to draw graphs and python to calculate statistics and place the results in a prebuilt html template.
• Analysis of Customer Data:-I used pandas, matplotlib, seaborn to derive insights from customer orders regarding distribution of orders across cities, payment methods, product categories, time and day of the week etc.
• Text Classification:-Given a text document we need classify it into one of the 9 classes. I used spacy for tokenization, lemmatization and removing stop words. I used scikit-learn for featurization and classification. I used CountVectorizer, TfidfVectorizer for featurization and MultinomialNB, KNeighboursClassifier, LogisticRegression, SVM, DecisionTreeClassifier, RandomForestClassifier, StackingClassifier, VotingClassifier for classification and reported log loss, precision and recall usinf confusion matrices..

Natural Language Processing related:-

• Sentiment Classification of Amazon Fine Foods reviews dataset
• Personalized Cancer Diagnosis Case Study

Computer Vision related:-I have implemented following algorithms using PyTorch from scratch as it helps to understand the underlying algorithms better.

• Drive CNN to predict steering angles
• Image inpainting using partial convolution
• Neural style transfer
• SRGAN for image super resolution
• Simple WDNet for watermark removal
• UNet for image segmentation
• YOLO-V1 for object detection