Lakshmi Jyothsna Machavarapu

A Data Science graduate skilled in PythOn, R, SQL, and visualization tools like Tableau and Power BI. Experienced in building predictive models, performing exploratory data analysis, and applying machine learning techniques to generate actionable insights. Completed projects such as fraud detection, housing price prediction, and global health analysis, showcasing a hands-on approach to solving real-world challenges. Recognized for adaptability, teamwork, and the ability to communicate complex data insights clearly to support decision-making.

4O

REAL/FAKE JOB POSTING PREDICTION (09/01/24 - 11/01/24): Developed a machine learning-driven system to detect fraudulent job postings, leveraging a Kaggle dataset comprising 18,000 job descriptions. Applied advanced Natural Language Processing (NLP) techniques, such as TF-IDF vectorization and lemmatization, and addressed class imbalance using SMOTE and ADASYN to enhance model performance for minority classes. Conducted extensive evaluation of models, including Random Forest, LSTM, SVM, and XGBoost, achieving 99% accuracy with SVM while maintaining a balance between precision and recall. Successfully deployed the final model as a Streamlit web application, enabling real-time fraud detection and improving safety and reliability on online job platforms.

BOSTON HOUSING PRICE PREDICTION (04/01/24 - 06/01/24): Utilized the Boston Housing dataset to predict median home values by applying advanced machine learning techniques. Explored and implemented various models, including Linear Regression, Random Forest, Gradient Boosting, and AdaBoost, with Random Forest achieving the highest performance . Performed comprehensive data preprocessing, feature engineering, and exploratory analysis to uncover key predictors such as socioeconomic status and average number of rooms per dwelling. Applied K-fold cross-validation to ensure model reliability and robustness, and delivered actionable insights to inform urban planning and real estate decision-making.

ADVANCED SALES ANALYTICS AND CUSTOMER SEGMENTATION PROJECT (09/01/23 - 11/01/23): Revolutionized sales analysis by steering a project that unveiled customer patterns through advanced clustering and introduced a savvy multi-class classification model. Infused demographic attributes for a richer model, ultimately delivering strategic insights for informed decision-making.

GLOBAL HEALTH INSIGHTS-MAPPING LIFE EXPECTANCY AND DETERMINENTS (04/01/23 - 06/01/23): Analyzed data from 193 countries over 15 years to explore how factors like health, economy, and social issues impact life expectancy. Employed a combination of map-based visuals, time series evaluations, and correlation analyses, alongside creating interactive dashboards for detailed data investigation, using statistical tools to identify and showcase trends. Identified critical socio-economic determinants impacting life expectancy globally, highlighting the significant role of healthcare expenditure and immunization coverage. Uncovered geographic patterns in disease prevalence and its correlation with mortality rates, emphasizing regions with urgent healthcare needs. Demonstrated the efficacy of data-driven approaches in public health policy planning, providing actionable insights for improving healthcare outcomes across different nations.

BRAIN TUMOR DETECTION AND CLASSIFICATION (01/01/22 - 04/01/22): The Proposed model for brain tumour detection and classification is SVM (Support Vector Machine). The main objective of SVM is to find a hyperplane in N-dimensional space that separates data points. It is the model used for both classification and regression. It tries to find out the best margin. It separates the classes and reduces the risk of error. It works with unstructured and semi-structured data like images and text. It is based on geometrical property of data. It is memory efficient and effective in High dimensional cases and reduces the risk of overfitting. 

• Programming for Everybody from University of Michigan (COURSERA)
• Python Programming (UDEMY)
• R programming (UDEMY)
• SQL for Beginners (UDEMY)
• Linear algebra for Data Science (UDEMY)
• AWS CERTIFIED (AWS)

BRAIN TUMOR DETECTION AND CLASSIFICATION,

International Journal of Innovative Research in Computer and Communication Engineering, 2320-9801, 2320-9798, 

www.ijircce.com, 8.165, 10, 5, 05/01/22

• TOPIC: Application of basic Python and libraries,
• CONDUCTED ON: 05/18/2020 - 05/20/2020,
• OUTCOMES: Login code with Tkinter
• TOPIC: Internet Of Things - The World Of Opportunities
• CONDUCTED ON: 05/25/2020 - 05/27/2020
• OCTCOMES: Identifying real time scenarios where IOT can provide better solutions

• Member of student activity council at VVIT.
• Head of alumni council in Student Activity Council at VVIT.
• Conducted alumni meet as a Head of alumni council for previous batches at VVIT in 2021