π I am passionate Data Analyst / Data Scientist π π who has worked in interationnal luxury retail group and equine hospital π’.
π I enjoy a lot in helping companies to discover and empower the potential of the data with data science tools and algorithms.
π Self-motivated and eager to learn different technologies and concepts. Recently, I am learning Scala and DBT.
- Sales and traffic forecasting
- Data anomaly detection
- Stock price prediction with algo-trading
- Mobile price range classification
- Recommendation system
- Social network analysis
- Sales performance dashboard
- Network traffic monitoring dashboard
- Drugs and vaccination usage dashboard
- Injuries and rehabilitation dashboard
- Python:
Numpy
,Pandas
,Scipy
,multiprocessing
- R:
ggplot2
,tidyr
- SQL
- Scala
- Java
- C++
- Software:
Power BI
,Tableau
,Snowflake
- Library:
Plotly
,Seaborn
,Matplotlib
- Frameworks:
statsmodels
,scikit-learn
- Regressions:
Linear
,Logistic
,Lasso
,Ridge
- Boosting and Trees:
XGBoost
,Catboost
,Adaboost
,Decision Tree
,Random Forest
- Clustering:
SVM
,K-Means
,LOF
,DBSCAN
- Time Series:
Prophet
,DeepAR
,SARIMAX
- Frameworks:
PyTorch
&Keras
- Models:
CNN
,RNN
,LSTM
,Transformer
- Large Language Model:
Hugging Face
- H2O.ai
- DBT
- Git
- Docker
- Airflow
- Platforms:
Google Cloud Platform
,Amazon Web Services
- Frameworks:
Spark
,Hadoop
- Python API:
PySpark
,Dask
- Regression Analysis
- Correlation Analysis
- Statistical Analysis
- Agile Methodology & Kanban
- Scrum & Sprint
- JIRA
- Confluence
- Streamlit
- Flask
- Data Cleaning
- Exploratory Data Analysis
- Feature Engineering
- Feature Extraction
- Prompt Engineering
- Zero-shot Inference
- One-shot Inference
- Few-shot Inference
- Fine-Tuning;
- Instruction Fine-Tuning
- LoRA
- Soft Prompt
- Utilized ensemble learning and deep learning to predict the price range of a used car in the North American market.
- Aimed to provide a data-driven price prediction for buyers and sellers to improve market efficiency. Try it out now!
- Applied
Logistic Regression
,gradient boosting
,random forest
,KNN
andNaive Bayes
to predict mobile phone price ranges. - Predict the price range of mobile phones based on their functionality and hardware components.
- Dataset from Kaggle and achieve 93.8% of weighted F1 score in baseline model -
Logistic Regression
.
- Perform predictive analysis and linear & logistic regression to predict the presence of West Nile Virus.
- Perform exploratory data analysis and data cleaning before data modelling
- Applied
SVC
,XGBoost
,Catboost
,Prophet
andCNN-LSTM
to predict stock prices. - Make trading decisions based on our model prediction and users' risk classification.
- Model Performance outperforms βBuy-and-holdβ Strategies.
- Perform network analysis to spot the key opinion leader in the network.
- Applied
Random Walk Generator
to extract information from local and global networks. - Applied
DeepWalk
andNode2Vec
for embedding stage - The
AUC-ROC
score reached 0.9323.
- Applied the
Neural Collaborative Filtering
(NCF) model in the recommendation system to predict the user's rating (1-6). - Used
Wide & Deep Learning
model for prediction. RMSE
dropped to 0.99.
- Applied
MLP
,Flair
,CNN
andBERT
with thePytorch
framework to predict the score of restaurant reviews. - Utilized various techniques such as
tokenization
,stopword removal
,stemming
andword embedding
withWord2Vec
&GloVe
- Perform data analysis with Tableau on Air Traffic Data
- Analyse the performance of major US Airlines such as Delta Airlines, American Airlines and Southwest Airlines
- Provide data-driven recommendations based on the past 10 years of Kickstarter campaign data.
- Perform visualizations to support the business insights and recommendations.
Sport | Stratrgic Games |
---|---|
Volleyball π | Texas Poker π |
Badminton πΈ | Mahjong π |
Bowling π³ |