kyoto-cheng Goto Github PK

🐱 My GitHub Data

📦 117.8 kB Used in GitHub's Storage

🏆 0 Contributions in the Year 2024

🚫 Not Opted to Hire

📜 23 Public Repositories

🔑 0 Private Repositories

I'm a Night 🦉

🌞 Morning                25 commits          ███░░░░░░░░░░░░░░░░░░░░░░   12.02 % 
🌆 Daytime                65 commits          ████████░░░░░░░░░░░░░░░░░   31.25 % 
🌃 Evening                103 commits         ████████████░░░░░░░░░░░░░   49.52 % 
🌙 Night                  15 commits          ██░░░░░░░░░░░░░░░░░░░░░░░   07.21 % 

📅 I'm Most Productive on Friday

Monday                   40 commits          █████░░░░░░░░░░░░░░░░░░░░   19.23 % 
Tuesday                  37 commits          ████░░░░░░░░░░░░░░░░░░░░░   17.79 % 
Wednesday                12 commits          █░░░░░░░░░░░░░░░░░░░░░░░░   05.77 % 
Thursday                 18 commits          ██░░░░░░░░░░░░░░░░░░░░░░░   08.65 % 
Friday                   70 commits          ████████░░░░░░░░░░░░░░░░░   33.65 % 
Saturday                 13 commits          ██░░░░░░░░░░░░░░░░░░░░░░░   06.25 % 
Sunday                   18 commits          ██░░░░░░░░░░░░░░░░░░░░░░░   08.65 % 

📊 This Week I Spent My Time On

🕑︎ Time Zone: Europe/London

🔥 Editors: 
No Activity Tracked This Week

💻 Operating System: 
No Activity Tracked This Week

Last Updated on 09/06/2024 01:18:22 UTC

• ML Web App for Stock Market Prediction From Daily News With Streamlit and Python
• Design A Dynamic Data Dashboard with Tableau
• Visualize Decrease in Arctic Sea Ice Extent with Flourish
• Generate Synthetic Dataset with the Synthetic Data Vault (SDV)
• NLP on Data Science Interview Questions & Answers Dataset
• Glassdoor Data Science Job Salary Predictor with Machine Learning
• Custom SpaCy Model Deployment with AWS ECR, SageMaker, and Lambda
• Overview of Building a Merchant Name Cleaning Engine with SequenceMatcher and spaCy
• Quantum Physics Word Cloud from 1994 to 2009

• Udacity SQL Nanodegree
  • Project-1: Created a report for the executive team on the topic of global & country-level deforestation by writing complex and advanced SQL queries.
  • Project-2: Investigated a poorly designed database and re-designed a new, normalised and performant database and migrated over data from the previous database.
• Udacity Data Visualisation Nanodegree
  • Project-1: Designed a Tableau dashboard which shows YoY performance in customer segmentation with different KPIs (Revenue, Profit, Profit ratio, etc).
  • Project-2: Created an animated data story and added an audio track for it to become a narrated Flourish story.
• Udacity NLP Nanodegree
  • Project-1: Built a spam email classifier using Naive Bayes algorithm that achieves Accuracy score: 0.988, Precision score: 0.972, Recall score: 0.941.
  • Project-2: Built a hidden Markov model for part of speech tagging with a universal tagset that achieves a test accuracy of 95.95%.
  • Project-3: Built a deep neural network end-to-end machine translation pipeline that accepts English text as input and return the French translation.
• Udacity Data Scientist Nanodegree
  • Project-1: Created a machine learning pipeline that categorises real messages sent during disaster events into 36 categories that achieves an overall accuracy ~90%.
  • Project-2: Collected and analysed a real user-article interaction dataset from IBM Watson Studio Platform and improves the user-article recommendation rate to ~95% with Final Matrix Factorization algorithm.
• Udacity Machine Learning Engineer Nanodegree
  • Project-1: Trained and deployed an XGBoost model on AWS SageMaker for IMDB movie review sentiment classification which meets ~85% accuracy.
  • Project-2: Built a plagiarism detector that examines a text file and performs binary classification by comparing containment and longest common subsequence with an accuracy of 96%.
  • Project-3: Deployed a machine learning web app on Streamlit to predict the stock market sentiment with over ~2000 daily news headlines and reached a ~84% accuracy (fine-tuned random forest classifier) & 81% (baseline logistic regression model).
• Udacity Deep Learning Nanodegree
  • Project-1: Built a simple neural network and used it to predict daily bike rental ridership on a timeseries dataset.
  • Project-2: Created a CNN to Classify Dog Breeds (using Transfer Learning VGG16) with an accuracy of 84%.
  • Project-3: Built a neural network (RNN & LSTM) that will generate a new ,"fake" TV script, based on patterns it recognizes in the training data.
  • Project-4: Defined and trained a DCGAN on a dataset of faces to get a generator network and generate new images of faces that look as realistic as possible.