Skip to content

This repository showcases 10 diverse data science and machine learning projects, ranging from classification tasks, time series prediction, and recommendation systems, to neural networks for handwritten recognition and mood detection. These projects highlight my proficiency with machine learning, deep learning, computer vision, tableau

Notifications You must be signed in to change notification settings

Honey28Git/LGMVIP-DataScience-Projects

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

47 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

LetsGrowMore Data Science Internship Projects

Project Portfolio Overview

This repository showcases 10 diverse data science and machine learning projects, ranging from classification tasks, time series prediction, and recommendation systems, to neural networks for handwritten recognition and mood detection. These projects highlight my proficiency with machine learning, deep learning, computer vision, and data analysis.

Projects Summary

  1. Iris Classification : Overview: Applied machine learning models to classify the Iris species using the Iris dataset. Approach: Used decision trees and other classification techniques. Conclusion: Achieved high classification accuracy by training a decision tree classifier on the well-known Iris dataset.

  2. Stacked LSTM for Price Prediction : Overview: Predicted stock prices for the next 28 days using the NSE Tata Global dataset. Approach: Developed a stacked LSTM model to capture the temporal patterns in stock price data. Conclusion: Successfully predicted the stock trend over the next month with satisfactory accuracy, providing insights into market movement.

  3. Music Recommendation System: Overview: Built a recommendation model based on the KKBOX Music Rec Challenge dataset to predict if a song will be played on repeat Approach: Trained a classification model to identify the top features influencing repeat song plays. Conclusion: Identified key features that influence music preferences, providing valuable insights into song recommendation mechanisms.

  4. Image to Pencil Sketch Conversion: Overview: Developed an OpenCV model to convert images into pencil sketches. Approach: Utilized image processing techniques in OpenCV to transform color images into artistic sketches. Conclusion: Successfully converted images to pencil sketches, demonstrating proficiency in computer vision.

  5. Global Terrorism Dataset EDA: Overview: Conducted exploratory data analysis (EDA) on the Global Terrorism dataset using Tableau, analyzing nearly 175 features. Approach: Visualized insights using Tableau for better understanding of global terrorism patterns. Conclusion: Provided a comprehensive analysis on terrorism attacks, accessible through my public Tableau dashboard: https://public.tableau.com/app/profile/honey.sam/viz/GlobalTerrorismAttackAnalysisWorkbook/AttacksSummary

  6. Decision Tree Classifier on Iris: Overview: Built a Decision Tree Classifier on the Iris dataset for species classification. Approach: Leveraged decision tree algorithms to classify the data. Conclusion: Achieved optimal classification performance, further confirming the effectiveness of tree-based models.

  7. Next Word Prediction Model: Overview: Created a neural network model for predicting the next word in a sequence using a text dataset. Approach: Trained an RNN-based model using Keras and TensorFlow on text data in .txt format, achieving a 72% accuracy. Conclusion: The model successfully predicts the next word in a sequence, showcasing pattern recognition in text data.

  8. MNIST Handwritten Equation Solver: Overview: Developed a CNN model to recognize handwritten numbers and symbols, evaluating simple equations. Approach: Used a CNN to read and interpret images of handwritten equations and output results. Conclusion: Accurately predicted and solved basic equations from handwritten input, automating equation solving.

  9. Neural Network for Equation Solving: Overview: Created a neural network model to solve handwritten equations by reading and interpreting the digits and symbols from images. Approach: Extended the CNN model to recognize equations and calculate the results. Conclusion: Successfully solved mathematical equations from images, combining computer vision and deep learning.

  10. Mood Detection and Song Suggestion: Overview: Built a CNN-based mood detection system to suggest songs based on facial expressions. Approach: Trained a CNN model on facial emotion data and achieved 74% accuracy in mood classification. Conclusion: The model can detect emotions and suggest music, showcasing the potential for emotion-based recommendation systems.

Conclusion

This repository highlights my ability to work with a wide variety of datasets and techniques, showcasing proficiency in machine learning, deep learning, computer vision, and data visualization. Each project has been carefully structured, from data preprocessing and model selection to performance evaluation.

About

This repository showcases 10 diverse data science and machine learning projects, ranging from classification tasks, time series prediction, and recommendation systems, to neural networks for handwritten recognition and mood detection. These projects highlight my proficiency with machine learning, deep learning, computer vision, tableau

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published