Lorem Ipsum
Python Data Preprocessing Using Pandas DataFrame, Spark DataFrame, and Koalas DataFrame
Preparing data for machine learning in Python See publication
See Publication
Object-Oriented Machine Learning Pipeline with mlflow for Pandas and Koalas DataFrames
End-to-end process of developing Spark-enabled machine learning pipeline in Python using Pandas, Koalas, scikit-learn, and mlflow. See publication
See publication
Automatic Machine Learning in Fraud Detection Using H2O AutoML
Machine Learning Automation in Finance. See publication
See publication
Deep Learning in Winonsin Breast Cancer Diagnosis
A deep learning approach for healthcare. See publication
See publication
Deep Learning for Natural Language Processing Using word2vec-keras
A deep learning approach for NLP by combining Word2Vec with Keras LSTM. See publication
See publication
Deep Multi-Input Models Transfer Learning for Image and Word Tag Recognition
A multi-models deep learning approach for image and text understanding. See publication
See publication
Deep Clustering for Financial Market Segmentation
A unsupervised deep learning approach for credit card customer clustering. See publication
See publication
Deep Learning for Image Classification on Mobile Devices
Mobile Image Classification App Development using Expo, React-Native, TensorFlow.js, and MobileNet. See publication
See publication
Deep Learning for Detecting Objects in an Image on Mobile Devices
Mobile Objects Detection App Development using Expo, React-Native, TensorFlow.js, and COCO-SSD. See publication
See publication
Deep Learning for Natural Language Processing on Mobile Devices
Reading Comprehension using Expo, React-Native, TensorFlow.js, and MobileBERT. See publication
See publication
Running Spark NLP in Docker Container for Named Entity Recognition and Other NLP Features
Using Spark NLP with Jupyter notebook for natural language processing in Docker environment. See publication
See publication
Common Time Series Data Analysis Methods and Forecasting Models in Python
Analyzing time series data for forecasting using ARIMA and LSTM models. See publication
See publication
Probabilistic Programming and Bayesian Inference for Time Series Analysis and Forecasting in Python
A Bayesian Method for Time Series Data Analysis and Forecasting in Python. See publication
See publication
Machine Learning for Building Recommender System in Python
Building Recommendation System Using Model-Based Collaborative Filtering in Python. See publication
See publication
Blindtext
Yuefeng, PhD in computer science, is a manager, senior data scientist at McDonald's. He
Lead the research & development of multiple computer vision projects using various advanced deep learning models for object detection, instance segmentation, object tracking, and object re-identification with one camera or across multi-cameras, including:
Food order accuracy intelligence with one camera: automatically detect food items in a customer food tray via deep learning object (e.g., food item) detection, and check customer order for any missing and/or wrong food item.
Deep learning food order tracking across multi-cameras via food items detection and tracking all of the food items in a food tray as a cluster of points.
Work utilization measurement intelligence via deep learning person detection, deep learning pose estimation, and innovative method of detecting active work time of each kitchen device (e.g., grill machine)
Personalized work utilization measurement automation via deep learning person detection, deep learning person tracking, deep learning pose estimation, attention-based deep person re-identification, and creative detection of active work time for each of the individual crew members.
Driving-through food order automation across multi-cameras via vehicle detection, re-identification, and tracking across multi-cameras using deep learning vehicle detection, deep learning vehicle tracking, and attention-based deep learning vehicle re-identification.
Serviced as the leading author of a well received white paper on computer vision for personalized work utilization
Contributed to trending analytics for foresight marketing intelligence
Yuefeng was a senior data scientist at Wavicle Data Solutions. He
Developed new deep learning computer vision model for fast food item detection using AWS SageMaker, Open CV, and YOLO v5
Developed new NLP (Natural Language Processing) model for sentiment analysis from customer feedback
Helped clients to enhance machine learning models with 10X CPU performance improvement
Developed clinical data statistical analysis and report generation software using AWS SageMaker with R kernel
Provided end-to-end machine learning services to clients
Authored machine learning project proposal
Utilized industry-leading AI platforms and frameworks such as GCP AI Platform, Databricks in AWS, AWS SageMaker, Hugging Face, Tensorflow, PyTorch
Contributed white papers on machine learning and deep learning applications
Played leading role in the establishment of a new data science team
Mentored junior data scientists and data engineers
Yuefeng was a senior data scientist at SMS Assist. He
Lead the facilities service pricing data analytics of residential and commercial reactive work orders
Developed a brand new data preprocessing and features engineering pipeline for the prediction of facilities service fair market price
Developed a brand new supervised machine learning system for the prediction of facilities service fair market price, including the work order affiliate invoice total, labor unit and total cost, materials total cost, travel total cost, freight total cost, and other cost
Established an industry standard machine learning process based on CRISP-DM (Cross-Industry Standard Process for Data Mining)
Guided and worked collaboratively with data engineering and DevOps teams to establish a sharable cloud machine learning environment in AWS using JupyterHub, Jupyter notebook, GitLab, S3, etc.
Developed machine learning and deep learning roadmap to management for the automation of residential and commercial reactive work order workflow
Identified and documented potential machine learning and deep learning opportunities
Yuefeng was a senior data engineer and principal associate at Capital One. He
Lead the architecture and development of the logistic regression and Random Forest machine learning engines for the prediction of customer profile matching, achieving more accurate prediction of performance than the legacy system developed by IBM
Lead the architecture, design, and development of multi-model machine learning system that can run multiple models concurrently for the prediction of probabilistic matching of different types of customer profiles with different features from different countries
Lead the research of innovative probabilistic customer matching methods via face recognition using deep learning algorithms such as convolutional neural networks (ConvNets)
Advised and collaborated with Capital One Card organization in the development of anomaly detection and forecasting system using deep learning algorithms such as deep autoencoder and recurrent neural network LSTM
Developed open source-based customer profile duplicate management system and presented it in Capital One 2019 Hackathon competition
Played leading role in the architecture and development of large-scale real-time data pre- processing, data blocking, features engineering, and model monitoring metrics calculation
Lead the architecture and development of large-scale real-time customer core analytical data streaming pipeline using Java, Python, Spark, Kafka, Spring Framework, and column-oriented distributed database
Assisted in teaching Capital One official machine learning training course, Chicago, 2019
Yuefeng was a senior system and application engineer at UniqueSoft, LLC. He
Served as project and team leader. Duties included project planning, management, and implementation of whole software development lifecycle for each project: System requirements solicitation, analysis, architecture/design, implementation, testing, and deployment.
Successfully developed several large-scale reverse-engineering/software-data-mining projects that utilize unsupervised machine learning algorithms to automatically extract application knowledge from legacy source code, including domain-specific application system features, architecture elements (architecture diagrams, call/control flow charts, dependency tables, sequence diagrams, etc.), complexity metrics (size, cyclomatic complexity, module dependency complexity, etc.), and major issues/improvement opportunities (e.g., replicated code detection and elimination, dead code detection and elimination, etc.).
Successfully developed various types of software products in different domains (e.g., telecommunications, defense, transportation, enterprise, etc.) and languages (TDL, Java, C/C++, PL/SQL, Cobol, JavaScript, etc.) using modern software engineering methodologies (e.g., software modernization via data mining, UML modeling and automatic code generation, Object- Oriented development, Service-Oriented development, Aspect-Oriented programming, Cloud computing, Agile processes, etc.), and Web application technologies such as J2EE, Spring Framework, Web services (SOAP, WSDL, etc.), JavaScript, Python, etc.
Yuefeng was a Distinguished Member of Technical Staff at Motorola Solutions. He
Worked collaboratively with senior management to develop strategy and approach to establishing Model-Driven Development (MDD) methodology and Agile development process for the development of large-scale parallel and distributed real time telecommunication systems.
Established partnerships with product, engineering, software automation teams, and MDD tools vendors such as IBM.
Developed and managed Motorola iDEN division-wide MDD roadmap, including software development processes, coding and design standards, protocols, platforms, programming languages, tools, etc.
Developed materials and conducted training for both technical and business colleagues.
Yuefeng authored book chapter and many peer-reviewed articles on machine learning, deep learning, data science, digital image processing, computer vision, etc. More detailed information about Yuefeng is available in
LinkedIn.
