ying zhang
Global Data Science Director at Code Worldwide- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
Topline Score
Bio
Credentials
-
MCP
Microsoft -
MCSE
Microsoft
Experience
-
Code Worldwide
-
United Kingdom
-
Advertising Services
-
1 - 100 Employee
-
Global Data Science Director
-
Aug 2021 - Present
Responsible for business development, scopes, designs, and implements ML models to support different clients, as well as the creation of the roadmap and growth of the data science function. Project examples include Recommendation, customer lifetime value, churn, promotion and price elasticity, content intelligence, path to loyalty etc for a high-end fashion client. Responsible for business development, scopes, designs, and implements ML models to support different clients, as well as the creation of the roadmap and growth of the data science function. Project examples include Recommendation, customer lifetime value, churn, promotion and price elasticity, content intelligence, path to loyalty etc for a high-end fashion client.
-
-
-
Infinity Sci.
-
France
-
Real Estate
-
Director
-
Apr 2018 - Aug 2021
For this boutique data analytics consulting company, worked with senior stakeholders to define how the business can create additional value through the utilization of data analytics • For a major real estate client, developed the forecast model to predict the potential time needed to fill commercial property for property developers • Worked on commercial automated valuation models (AVM) to combine location, macroeconomic, costs, and service data even with residential property price and… Show more For this boutique data analytics consulting company, worked with senior stakeholders to define how the business can create additional value through the utilization of data analytics • For a major real estate client, developed the forecast model to predict the potential time needed to fill commercial property for property developers • Worked on commercial automated valuation models (AVM) to combine location, macroeconomic, costs, and service data even with residential property price and stock market info to predict commercial property’s current or future value. • Applied Text classification to automatically assign tax codes to tax documents, reduce manual tax assignment process time from months to minutes • Applied feature importance to identify potential drivers for the London property market • Applied density-based clustering to segment clients with follow-up strategies to improve property investment clients care service • For a major insurance client, developed cross-sell and reward solution for premium customers and long tail short head recommendation with multi-class boosting • For a retail research client, identified important features which may affect customer purchase behaviours, and applied Time Series Forecasting to predict sales for each store, with holiday and promotion data • For a global banking client, designed analytics solution for market forecasting and cross-sell with segment tailored strategy and proposition Show less For this boutique data analytics consulting company, worked with senior stakeholders to define how the business can create additional value through the utilization of data analytics • For a major real estate client, developed the forecast model to predict the potential time needed to fill commercial property for property developers • Worked on commercial automated valuation models (AVM) to combine location, macroeconomic, costs, and service data even with residential property price and… Show more For this boutique data analytics consulting company, worked with senior stakeholders to define how the business can create additional value through the utilization of data analytics • For a major real estate client, developed the forecast model to predict the potential time needed to fill commercial property for property developers • Worked on commercial automated valuation models (AVM) to combine location, macroeconomic, costs, and service data even with residential property price and stock market info to predict commercial property’s current or future value. • Applied Text classification to automatically assign tax codes to tax documents, reduce manual tax assignment process time from months to minutes • Applied feature importance to identify potential drivers for the London property market • Applied density-based clustering to segment clients with follow-up strategies to improve property investment clients care service • For a major insurance client, developed cross-sell and reward solution for premium customers and long tail short head recommendation with multi-class boosting • For a retail research client, identified important features which may affect customer purchase behaviours, and applied Time Series Forecasting to predict sales for each store, with holiday and promotion data • For a global banking client, designed analytics solution for market forecasting and cross-sell with segment tailored strategy and proposition Show less
-
-
-
McKinsey & Company
-
Business Consulting and Services
-
700 & Above Employee
-
Data scientist
-
Mar 2015 - Apr 2018
Worked on advanced data analytics for customer journey analytics solution team, applied sequence data mining technology to understand and predict sequential data patterns. Engagements included: • For a major insurance developed a Customer churn model to reduce customer attrition • For a major telecom client, aggregated and analyzed operational data by customer journey across multiple channels. Identified underlying pain points and optimized customer journey • For a global banking… Show more Worked on advanced data analytics for customer journey analytics solution team, applied sequence data mining technology to understand and predict sequential data patterns. Engagements included: • For a major insurance developed a Customer churn model to reduce customer attrition • For a major telecom client, aggregated and analyzed operational data by customer journey across multiple channels. Identified underlying pain points and optimized customer journey • For a global banking client drove digital engagement of financially active customers with tailored multichannel tactics • For a major shipping client applied position and schedule data to predict vessel arrivals, helping port players improve planning operations and increase asset utilization • Employed tools that included: Hadoop, AWS, Hive, R, Python, Tableau, Random Forest, Generalized Boosted Regression, Extreme Gradient Boosting Show less Worked on advanced data analytics for customer journey analytics solution team, applied sequence data mining technology to understand and predict sequential data patterns. Engagements included: • For a major insurance developed a Customer churn model to reduce customer attrition • For a major telecom client, aggregated and analyzed operational data by customer journey across multiple channels. Identified underlying pain points and optimized customer journey • For a global banking… Show more Worked on advanced data analytics for customer journey analytics solution team, applied sequence data mining technology to understand and predict sequential data patterns. Engagements included: • For a major insurance developed a Customer churn model to reduce customer attrition • For a major telecom client, aggregated and analyzed operational data by customer journey across multiple channels. Identified underlying pain points and optimized customer journey • For a global banking client drove digital engagement of financially active customers with tailored multichannel tactics • For a major shipping client applied position and schedule data to predict vessel arrivals, helping port players improve planning operations and increase asset utilization • Employed tools that included: Hadoop, AWS, Hive, R, Python, Tableau, Random Forest, Generalized Boosted Regression, Extreme Gradient Boosting Show less
-
-
-
AlixPartners
-
United States
-
Business Consulting and Services
-
700 & Above Employee
-
Associate
-
Jun 2014 - Feb 2015
For a major retail bank worked on Data Integration, Customer Insight and Reporting. Also applied Text sentimental Analysis for customer feedback text classification. (Python NLTK) For a major retail bank worked on Data Integration, Customer Insight and Reporting. Also applied Text sentimental Analysis for customer feedback text classification. (Python NLTK)
-
-
-
Expedia
-
Singapore
-
Internet Publishing
-
100 - 200 Employee
-
Online Data Science Manager
-
Jun 2013 - Apr 2014
• Collaborative filtering recommendation engine design for up-sell and cross-sell • Content based recommendation and basket content recommendation design • Search engine marketing(SEM) keywords folding and keywords hierarchy • Omniture hit data analysis, multivariate testing, uplift modelling • Collaborative filtering recommendation engine design for up-sell and cross-sell • Content based recommendation and basket content recommendation design • Search engine marketing(SEM) keywords folding and keywords hierarchy • Omniture hit data analysis, multivariate testing, uplift modelling
-
-
-
Dunnhumby
-
United Kingdom
-
Advertising Services
-
700 & Above Employee
-
Senior Data Scientist
-
Dec 2012 - May 2013
Design and improve personalization models such as complement and substitute recommendation base on customer lifestyle segmentation, shopping behavior, click stream data, location, weather etc. Design and improve personalization models such as complement and substitute recommendation base on customer lifestyle segmentation, shopping behavior, click stream data, location, weather etc.
-
-
-
-
Data Science Manager
-
Sep 2010 - Nov 2012
My job involves mainly project consulting, technical support and some development as well. • Ladbrokes cross sell project, recommend best product for customer at best time • McDonald’s up sell project, design the content of bundle meal and store segmentation • Credit Risk Scorecard design include Data Acquisition and Integration, Data Profiling, Variable Selection, Logistic Regression, Coarse Classing and WOE, Model Evaluation, Scorecard Scaling and Deployment (SAS). • Scorecard… Show more My job involves mainly project consulting, technical support and some development as well. • Ladbrokes cross sell project, recommend best product for customer at best time • McDonald’s up sell project, design the content of bundle meal and store segmentation • Credit Risk Scorecard design include Data Acquisition and Integration, Data Profiling, Variable Selection, Logistic Regression, Coarse Classing and WOE, Model Evaluation, Scorecard Scaling and Deployment (SAS). • Scorecard Reject Inference implementation, including Simple Augmentation, Fuzzy Augmentation, Parceling • Uplift modeling by using both the treated and control customers to build the predictive model that focuses on the incremental response. • Text sentimental Analysis (Lexalytics) • AXA customer feedback text classification. (Python NLTK) • TalkTalk customer churn model to reduce customer attrition • Honda UK customer Life time value project to identify the potential customer when and what type of car they would be interested. (Decision tree, Logistic Regression) Show less My job involves mainly project consulting, technical support and some development as well. • Ladbrokes cross sell project, recommend best product for customer at best time • McDonald’s up sell project, design the content of bundle meal and store segmentation • Credit Risk Scorecard design include Data Acquisition and Integration, Data Profiling, Variable Selection, Logistic Regression, Coarse Classing and WOE, Model Evaluation, Scorecard Scaling and Deployment (SAS). • Scorecard… Show more My job involves mainly project consulting, technical support and some development as well. • Ladbrokes cross sell project, recommend best product for customer at best time • McDonald’s up sell project, design the content of bundle meal and store segmentation • Credit Risk Scorecard design include Data Acquisition and Integration, Data Profiling, Variable Selection, Logistic Regression, Coarse Classing and WOE, Model Evaluation, Scorecard Scaling and Deployment (SAS). • Scorecard Reject Inference implementation, including Simple Augmentation, Fuzzy Augmentation, Parceling • Uplift modeling by using both the treated and control customers to build the predictive model that focuses on the incremental response. • Text sentimental Analysis (Lexalytics) • AXA customer feedback text classification. (Python NLTK) • TalkTalk customer churn model to reduce customer attrition • Honda UK customer Life time value project to identify the potential customer when and what type of car they would be interested. (Decision tree, Logistic Regression) Show less
-
-
-
-
Principal Data Scientist
-
Feb 2010 - Aug 2010
analyze huge amount of transactional and behavioral data, and apply probabilistic, machine learning techniques to build models to improve targeting and revenue. • Time series data analysis, I used R and Java to develop outlier detector for all the campaigns, networks, advertisers performance. All the data has time series with seasonal pattern. I apply ARIMA and LOESS model to find the outliers and it can automatically send email to related teams once anomaly is detected • Genetic… Show more analyze huge amount of transactional and behavioral data, and apply probabilistic, machine learning techniques to build models to improve targeting and revenue. • Time series data analysis, I used R and Java to develop outlier detector for all the campaigns, networks, advertisers performance. All the data has time series with seasonal pattern. I apply ARIMA and LOESS model to find the outliers and it can automatically send email to related teams once anomaly is detected • Genetic algorithm for multi-objectives optimization of rule engine. There are different business goals we need to achieve at same times, these goals may depends on each other, optimize one may damage another. I use Python and MS SQL to develop a tool which can automatically generate a set of rules to achieve the maximum profit, sales, revenue and ROI • Real time online bidding, in order to achieve the best ROI, profit, we cluster products and users in different groups, then based on the history bidding prices to predict the real time accurate bidding prices for the impressions. Show less analyze huge amount of transactional and behavioral data, and apply probabilistic, machine learning techniques to build models to improve targeting and revenue. • Time series data analysis, I used R and Java to develop outlier detector for all the campaigns, networks, advertisers performance. All the data has time series with seasonal pattern. I apply ARIMA and LOESS model to find the outliers and it can automatically send email to related teams once anomaly is detected • Genetic… Show more analyze huge amount of transactional and behavioral data, and apply probabilistic, machine learning techniques to build models to improve targeting and revenue. • Time series data analysis, I used R and Java to develop outlier detector for all the campaigns, networks, advertisers performance. All the data has time series with seasonal pattern. I apply ARIMA and LOESS model to find the outliers and it can automatically send email to related teams once anomaly is detected • Genetic algorithm for multi-objectives optimization of rule engine. There are different business goals we need to achieve at same times, these goals may depends on each other, optimize one may damage another. I use Python and MS SQL to develop a tool which can automatically generate a set of rules to achieve the maximum profit, sales, revenue and ROI • Real time online bidding, in order to achieve the best ROI, profit, we cluster products and users in different groups, then based on the history bidding prices to predict the real time accurate bidding prices for the impressions. Show less
-
-
-
Middlesex University
-
United Kingdom
-
Higher Education
-
700 & Above Employee
-
Visiting Lecturer
-
Sep 2009 - Jan 2010
As Visiting Lecturer and Module leader of “Data Warehouse & Business Intelligence”, I taught about 150 final year students, the module focuses on “business aspects” of data mining and data warehousing, including hands-on lab activities. As Visiting Lecturer and Module leader of “Data Warehouse & Business Intelligence”, I taught about 150 final year students, the module focuses on “business aspects” of data mining and data warehousing, including hands-on lab activities.
-
-
-
University of West London
-
United Kingdom
-
Higher Education
-
700 & Above Employee
-
Data Mining Engineer
-
Jun 2008 - Jan 2010
As customer retention is one of the most important problems for commercial companies, student retention is the key for university success. I joined this small team to applied Data Mining and natural language processing technology to evaluate course suitability, monitor student academic behavior. Students are clustered based on their profile and performance, and then different data mining models are built for each group to predict student performance progress and their probability to… Show more As customer retention is one of the most important problems for commercial companies, student retention is the key for university success. I joined this small team to applied Data Mining and natural language processing technology to evaluate course suitability, monitor student academic behavior. Students are clustered based on their profile and performance, and then different data mining models are built for each group to predict student performance progress and their probability to dropout. We build the University Data Warehouse, integrated data from 9 different sources then aggregate and analysis data by Data Cube. I also used Business Intelligence reporting tools to design dashboard report, email templates and according to data mining results to personalize the intervention for students to achieve better Student Retention. This project achieves great success and high praise from both university and funding organization. Some conference and journal paper are also published Show less As customer retention is one of the most important problems for commercial companies, student retention is the key for university success. I joined this small team to applied Data Mining and natural language processing technology to evaluate course suitability, monitor student academic behavior. Students are clustered based on their profile and performance, and then different data mining models are built for each group to predict student performance progress and their probability to… Show more As customer retention is one of the most important problems for commercial companies, student retention is the key for university success. I joined this small team to applied Data Mining and natural language processing technology to evaluate course suitability, monitor student academic behavior. Students are clustered based on their profile and performance, and then different data mining models are built for each group to predict student performance progress and their probability to dropout. We build the University Data Warehouse, integrated data from 9 different sources then aggregate and analysis data by Data Cube. I also used Business Intelligence reporting tools to design dashboard report, email templates and according to data mining results to personalize the intervention for students to achieve better Student Retention. This project achieves great success and high praise from both university and funding organization. Some conference and journal paper are also published Show less
-
-
Education
-
2004 - 2008
University of Surrey
PhD, Machine Learning, Decision Making -
2002 - 2004
Newcastle University
MPhil, Web / Text Mining
Community
