Learn how Corporate Functions like HR and Finance can leverage sophisticated solutions for predictive modeling, data mining and advanced visualization.
Follow me on twitter @hschliebs
Margin2years of history minimumProfitabilityCompany Performance: Correlations to external factors like GDP, Employment, S&P etcRegression (1 factor)Neural Network (all factors)Monte Carlo (Optimization)As many variables as possible!!!!Financial RiskWhere do we want to see outliers?Performance ManagementClustering / Payments:Late Payments (categorize clusters based on payment history: worst group, e.g.) as much variables as possible as inputProfitable ProductsCombine algorithms, e.g.ROB KUGEL VENTANA: DIFFERENTIATION: WHAT DO WE DO ABOUT IT (PROBLEMS) NOW??? AND: VISUALIZATION IN CONTEXT TO THE SOLVED QUESTIONS2 scenarios:Planning supportSpecific AnalysisApproach: something that’ll do the basics flawlessly:easier visualization, easier interaction (data sets) and combination of financial and operational data (absence of detail in operational data)Start with FP&A people / roles people who put together the financial plans. Describe it as a tool that makes the budgeting and planning forecast much more useful, FINANCE IS A STRATEGIC PARTNER FOR THE REST OF THE ORGANIZATION. 3 personas: PhD (work out model), Analyst (applies it to domain) and End User/Executives (uses results / visualizations and takes action)
Employee Performance Analysis: compare active vs. non-active performances per regionEffectiveness Analysis: Historical view on successful career paths – between job groupsWorkforce Performance: The bubblechartshowsustheold and thenewperformanceratings for theclusters - wecanseethatthe for cluser 3 (orange box) thenewratingsaresignificantlowerthantheoldratings – didthesepeopleleavebecausetheywere not successful?Proactive Talent turnover detection: Optimized career path management: build out profiles, i.e. clusterswith high densityallowingpredicthow a groupofemployeeswithcommoncharacteristicswouldbehave – thiswouldallowmitigatinginterventionsasactionscouldbetakenmorefocused, i.e. adresswithretention, development, mentoringspecificemployeegroups.Turnover Clustering: overviewoftheclustersthatcouldbeidentified, thesizetheyhave and theirdensity and wecanhave a high levelviewtocompareclusters
Employee Performance Analysis: compare active vs. non-active performances per regionEffectiveness Analysis: Historical view on successful career paths – between job groupsWorkforce Performance: The bubblechartshowsustheold and thenewperformanceratings for theclusters - wecanseethatthe for cluser 3 (orange box) thenewratingsaresignificantlowerthantheoldratings – didthesepeopleleavebecausetheywere not successful?Proactive Talent turnover detection: Optimized career path management: build out profiles, i.e. clusterswith high densityallowingpredicthow a groupofemployeeswithcommoncharacteristicswouldbehave – thiswouldallowmitigatinginterventionsasactionscouldbetakenmorefocused, i.e. adresswithretention, development, mentoringspecificemployeegroups.Turnover Clustering: overviewoftheclustersthatcouldbeidentified, thesizetheyhave and theirdensity and wecanhave a high levelviewtocompareclusters