Portfolio - Hark Poon

Python Project

EDA on Netflix movies & TV shows

The goal of this project is performing EDA on Netflix data to classify/group the Netflix shows into certain clusters such that the shows within a cluster are similar to each other and the shows in different clusters are dissimilar to each other.

View Project
on GitHUB

Project Struture

Imported Libraries

Project Details

The dataset contained about 7787 records, and 11 attributes.

1. Started by dealing with the dataset's missing values and doing exploratory data analysis (EDA).

2. Creating cluster using following attributes: director, cast, country, genre, rating and description. The values in these attributes were tokenized, preprocessed, and then vectorized using TFIDF vectorizer.

3. We used Principal Component Analysis (PCA) to handle the curse of dimensionality.

4. We built Two types of clusters using the K-Means Clustering and Agglomerative Hierarchical clustering algorithm and find out optimal number of clusters using different technique such as elbow method, silhoutte score and dendogram etc.

5. A content based recommender system was built using the similarity matrix obtained after using cosine similarity. This recommender system will make 10 recommendations to the user based on the type of show they watched.

Project video

PowerBI Project

Business Insight 360

Fast-growing consumer electronics company AtliQ Hardwares struggles with scattered Excel reports, hindering data-driven decisions. Project implementation on Power BI for business insights will unify data, reveal trends and empower smart choices for optimal operations and a competitive edge.

Home page

Complex Data Modeling involving 12+ tables.

Project Details

This project aims to design a single report with multi-view Power BI dashboard for AtliQ Hardware. This interactive tool will empower stakeholders across sales, finance, supply chain, marketing, and executive teams to understand business trends and make data-driven decisions to scale operations

Key areas of focus include:
1. Robust Data Modeling
2. User-empathetic report design
3. Drillable Insights

Here is the list of topic implemented on this report -
1. Power Query (Basic and Advanced operation)
2. Complex Data Modeling involving 12+ tables.
3. Basic and Advanced DAX functionalities
4. Choosing the right visual and formatting
5. Dashboard design principle
6. Page Navigation and Bookmarks
7. what-if parameter & field parameter
8. UI / UX design
9. Deployment in Power BI service
10. On premises Data Gateway (Personal mode)
11. Stakeholder feedback Implementation
12. KPI metrics for Finance, Sales, Supply chain, Marketing
etc. & many more....

Live Interactive Dashboard

Python Project

EDA on Google Play Store

The Play Store apps data has enormous potential to drive app-making businesses to success. Actionable insights can be drawn for developers to work on and capture the Android market. Explore and analyze the data to discover key factors responsible for app engagement and success.

View Project
on GitHUB

Libraries used

Project Structure

Project Details

This project is divided into two parts.

1. Data Preprocessing
- Analyze and explore the data, converting data type of columns from object to numerical or datetime format as per needed to make it ready for our analysis.
- Dropping irrelevant columns, removing null values and deleting the duplicate row which are present in dataframe.
- Imputing null values with Mean or Median for numerical attribute and Mode for categorical attribute.

2. Data Visualization
- Plotted several graphs and extracted more insights from our dataset which will help the developers in rolling out the app in the play store.

Hello, I am

Hark Poon

1

2

About

Technical

My

EDA on Netflix movies & TV shows

Business Insight 360

EDA on Google Play Store

Awards

Contact Me

Email

Github

Linkedin

Call

Hark Poon

Download Resume