.about()
# Education
Master’s in Data Science, Deep Learning specialization
Liverpool John Moore University
Expected Q1 2024
MicroMaster’s in Statistics and Data Science
Massachusetts Institute of Technology
Expected Q2 2024
Professional Program in Data Science and Machine Learning
Massachusetts Institute of Technology
Oct 2022
# Experience
Digital/Social Media Art Director
Leo Burnett Vietnam, Leo Burnett Malaysia
2019 – currently
– Developed strategies and executed digital/social media marketing campaigns for clients: Samsung, BMW, Quaker, Garnier.
– Collaborated with stakeholders and worked with crossfunctional teams to deliver under tight deadlines and pressure.
– Led a team of creatives in creating scroll-stopping contents and delivered expectation-exceeding KPI.
Freelancing Art Director
Publicis Vietnam, T&A Ogilvy, Dsquare Digital
2018 – 2019
Junior Art Director
TBWA\Group Vietnam
2015 – 2017
– Developed and produced effective advertising campaigns, from ideation to production and qualitative testing (consumer test) to ensure that the key message resonated with the target consumers.
– Clients: Vinamilk, Bosch, Acecook.
.projects()
Hand Gesture Recognition with CNN+RNN
- CNN
- RNN
- TensorFlow
- ComputerVision
Develop a model to recognize 5 different hand gestures from image sequence. Attempt to build the model from scratch and finetune retrained models. Final model is a finetuned MobileNet + LSTM due to its low-latency, low-power, parameterized to work under the resource constraints. Accuracy 100% on train, validation and test cases.
Natural Language Processing and Sentiment Analysis with Neural Network
- NLP
- ANN
- TensorFlow
- NLTK
- RegEx
- Steam API
- Unstructured data
Customer Experience prediction hackathon
- Hackathon
- Classification
- TensorFlow
- ANN
- Random Forest
- XGB
Built and fine-tuned a deep learning model to predict customer satisfaction from survey data during a 72-hour hackathon. Achieved an accuracy of 95.7% on the final model and arrived at final rank 3 on the leaderboard.
Natural Disaster Prediction / Prevention with convolution neural network
- Classification
- TensorFlow
- CNN
- Feature engineering
Developed a deep learning model to classify if a landslide occurred or not based on terrain data. Main challenges includes processing data with high dimensionality, multicollinearity, latent variables, and data imbalance.
Telecom Customer Churn Prediction and Retention Strategy
- Classification
- PCA
- Logistic Regression
- Random Forest
- ADABoost
- XGBoost
- SVM
Identifed the factors driving churns in high value customers and devise strategies to manage churners and developed model to predict churners, using Logistic Regression, Random Forest, AdaBoost, XGBoost and SVM. Handled large number of highly correlated variables and class imbalance using PCA, upsampling and deriving new features.
Lead Conversion in online education platform
- Classification
- Logistic Regression
Built a predictive model to predict promising leads and improve lead conversion rate. Identified the important factors that drive leads and optimized models/strategies to adapt to different business requirements.
.awards()
# 3rd place, Hackathon
Machine Learning Hackathon by GreatLearning & MIT IDSS
Oct 2022
# 2nd place, Hackathon
Wilson Analytics Mega Hackathon in Machine Learning
Dec 2022
