南澳大学计算机科学（Causal Data Mining）Ph.D. 目前在澳大利亚税务局担任Assistant Director Data Scientist 。他在学术和工业领域的数据科学项目中拥有约十年的经验，深刻理解数据挖掘和分析的复杂性，致力于通过高级数据科学技术推动政府数据的透明度和效率。
In the highly competitive airline industry, Tigerair has always been committed to providing its passengers with an exceptional flying experience. However, with the ever-changing market dynamics and the increasing diversity in passenger needs, Tigerair is faced with the challenge of enhancing passenger satisfaction, strengthening brand loyalty, and increasing market share. Despite various measures already in place, the company recognizes that to maintain a leading position in the fierce market competition, it needs to deeply understand the satisfaction levels of its passengers and the factors that influence them, and based on these insights, implement effective strategies.
项目内容
Launch your data science career in just 8 weeks. This comprehensive project delves into core machine learning principles and practical applications, encompassing key areas such as problem formulation, exploratory data analysis (EDA), preprocessing, feature engineering, and the development and evaluation of machine learning models. You will hone your development skills to execute each phase of the data science lifecycle, utilizing Python and Jupyter Notebook.
The curriculum also covers advanced techniques, data/ML pipeline, specifically focusing on the widely utilized pipeline employed by major artificial intelligence companies.
Beyond merely introducing these concepts, the project is designed to bolster your practical skills and real-world experience.
Guiding you through this immersive learning experience is Dr. Saisai Ma, the Data Science Director (acting) at the Australian Taxation Office (ATO). Dr. Ma is dedicated to providing hands-on guidance to help you forge your path in the data science field.
导师介绍
Saisai was awarded a Ph.D. degree in Computer Science (Causal Data Mining) at the University of South Australia. Currently he is an Assistant Director Data Scientist at Australian Taxation Office. He has around 10 years’ experience on data science projects, in both academic and industrial domains.
技术栈
1. Data Analysis & Preprocessing
Programming Languages: Python and Jupyter notebook
Libraries: Pandas for data manipulation, NumPy for numerical operations
Tech Stack: Handling missing values, dealing with duplicate records, managing data types, addressing inconsistencies
2. Data Visualization
Libraries: Matplotlib and Seaborn for Python
Techniques: Exploratory data analysis (EDA), visualizing variation and covariation among variables
3. Feature Engineering & Selection
Libraries: scikit-learn and feature-engine for Python
Techniques: Creating new features, selecting significant features, dimensionality reduction
4. Machine Learning
Concepts: Basic to advanced ML concepts, model selection
Libraries: scikit-learn for Python
Techniques: Implementing various ML models, bias-variance tradeoff understanding
5. Hyperparameter Tuning
Techniques: Grid search, random search, Bayesian optimization
Libraries: scikit-learn's GridSearchCV and RandomizedSearchCV for Python
6. Model Evaluation
Techniques: Cross-validation, understanding and applying various metrics (e.g., accuracy, precision, recall, F1 score)
7. Pipelines
Libraries: scikit-learn's Pipeline for Python
Components: Preprocessing pipeline, feature engineering pipeline, training pipeline, scoring pipeline
8. Understanding & Framing DS Problems
Skills: Translating business problems into data science problems, identifying key objectives and metrics
项目目标
Tigerair aims to invite students and professionals from different backgrounds to a data science boot camp, utilizing the company's extensive collection of flight and passenger satisfaction survey data to analyze and identify the key factors affecting passenger satisfaction. The goal of the project is to uncover the relationship between passenger satisfaction and various aspects of the service, and to predict which factors most significantly impact passenger satisfaction. Through this project, Tigerair hopes to:
Identify Key Satisfaction Factors: Determine which service attributes (e.g., ease of online booking, seat comfort, in-flight entertainment) have the most significant impact on passenger satisfaction.
Optimize Service Experience: Formulate specific improvement measures based on the analysis results to enhance the overall satisfaction of passengers with Tigerair flights.
Provide Personalized Services: Develop more personalized services by analyzing the needs of different passenger groups to meet their specific requirements.
Improve Operational Efficiency: Identify strategies to reduce delays and improve punctuality by analyzing the relationship between flight delays and passenger satisfaction, thereby enhancing both passenger satisfaction and company efficiency.
Customer satisfaction analysis/prediction is one of the most common problems in all kinds of companies, to better understand customer needs and improve service quality and customer stickiness. If you are seeking employment with a company that manufactures products or offers services, then this project is indispensable for you.
案例展示
指导方式
Groups of 4 people will receive 3 hours of project coaching per week. The project will be developed over a period of 8 weeks, with each week being a different task, following the pace of the tutor throughout the project.
