Welcome to My Tech-Journal

A space where I share my thoughts, experiences, and learnings on technology, programming, and innovation. Expect in-depth articles, tutorials, and projects that showcase my expertise in web development, cloud computing, machine learning, and more!

Check out the latest posts below!!


DAY 26- TIME SERIES FORECASTING OF ELECTRICITY CONSUMPTION USING LSTM (INTRO TO DEEP LEARNING)

On Day 26, I used LSTM (Long Short-Term Memory) to build a model for forecasting hourly electricity consumption. Using past 60 hours of consumption data as input, the model was able to predict the next hour’s consumption with high accuracy. After training, the model’s predictions closely followed the actual electricity demand in the test set, demonstrating the effectiveness of LSTM for time series forecasting.

DAY 25 - SENTIMENT ANALYSIS OF CUSTOMER REVIEWS USING TRADITIONAL NLP TECHNIQUES

This project marked our introduction to NLP, where we revisited familiar steps while gaining new insights into the specific requirements for NLP tasks. The experience was enriching, and I’m excited to apply this knowledge to upcoming challenges.

DAY 24 - K-MEANS CLUSTERING TO SEGMENT CUSTOMERS BASED ON BEHAVIOR

We built a customer segmentation model using K-Means clustering, selecting 4 clusters based on the Elbow Method. After preprocessing and scaling the features, the model successfully segmented customers based on their annual income and spending score, revealing distinct customer groups with similar behaviors. This segmentation can be used by businesses to personalize marketing strategies and enhance customer experience.

DAY 23 - FRAUD DETECTION IN FINANCIAL TRANSACTIONS USING LOGISTIC REGRESSION AND RANDOM FOREST

We built two models—Logistic Regression and Random Forest—to detect fraudulent financial transactions. While Logistic Regression achieved high accuracy, it struggled with recall, missing many fraud cases. Random Forest, on the other hand, performed significantly better, with improved precision and recall, making it more effective at identifying fraud in this highly imbalanced dataset.

DAY 22 - RECOMMENDER SYSTEM WITH MATRIX FACTORIZATION

It marks Day 22 of my 30 Days, 30 Machine Learning Projects Challenge. Today’s challenge was build Recommender System With Matrix Factorization. Curious about how it went? Read on to see the results!

DAY 21 - DEPLOY A MACHINE LEARNING MODEL USING FASTAPI AND HEROKU FOR REAL-TIME PREDICTIONS

It marks Day 21 of my 30 Days, 30 Machine Learning Projects Challenge. Today’s challenge was focused on deploying a machine learning model for real-time predictions. I chose to deploy my Fake News Detection model using FastAPI as the API framework and Heroku for hosting the app. Curious about how it went? Read on to see the results!

DAY 20 - 30 DAYS 30 ML PROJECTS: CREATE A TOPIC MODEL USING LATENT DIRICHLET ALLOCATION (LDA)

Today marks Day 20 of my 30 Days, 30 Machine Learning Projects Challenge. The problem is to Create a topic model using Latent Dirichlet Allocation (LDA). Curious about how it went? Read on to see the results!

DAY 19 - 30 DAYS 30 ML PROJECTS: CUSTOMER CHURN PREDICTION WITH XGBOOST

Today marks Day 19 of my 30 Days, 30 Machine Learning Projects Challenge. The problem is to Predict customer churn using XGBoost. Curious about how it went? Read on to see the results!

DAY 18 - 30 DAYS 30 ML PROJECTS: TIME SERIES FORECASTING OF STOCK PRICES WITH ARIMA MODEL

On Day 18, we used the ARIMA model to predict stock prices based on historical closing prices. By using the ARIMA (Auto-Regressive Integrated Moving Average) approach, we built a model that can predict short-term future values in the time series.

DAY 17 - 30 DAYS 30 ML PROJECTS: PREDICT DIABETES ONSET USING DECISION TREES AND RANDOM FORESTS

Hey, It is Day 17 of the 30 Days 30 Machine Learning Project Challenge, and today we predicted diabetes onset using Decision Trees and Random Forests. Curious about how it went? Read on to see the results!