MIcompany - Executive Data Analytics Programma

About MIcompany - Executive Data Analytics Programma

For companies seriously seeking to harvest the promised Data Analytics gold, investing in analysts is not enough. Executives play a pivotal role in turning great analytics into great top and bottom line results. Their willingness to transform the company to a data-driven organization and to adopt and use factual insights is crucial. Data Analytics knowledge of senior management greatly enhances putting great insights into action. The Executive Data Analytics (EDA) program is focused on Data Analytics business impact and uses the modelling process as guidance. In 2 interactive days theory is combined with practical case work producing models on relevant business topics. Participants take in strategic and key data analytics knowledge to master eight topics:

1.Use case identification and roadmap to transform into a data-driven company, the integrated method to establish automated predictive models in business processes and the way-of-working 2. Description of the expected relations between dependent and independent varaibles, mediators and moderators of the conceptual model 3. The data value chain, including the importance of feature data to enhance predictions, the need for an universal definition of variables, and structural and discovery data access routes. 4. Key concepts of descriptive analytics: granulartity of insights (modus median and distributions) their significance (p-value<0.05) and causality (use hypothesis and involve business knowlegde). 5.The modelling landscape, as defined by the Learning strategy (AI, ML, DL), learning method (supervised, unsupervised learning, reinforcement learning) and the model goal (regression, classification, clustering and dimensionality reduction) 6. The essence of modelling in R: linear regression, ARIMAX, logit regression, (CART) decision trees, random forest, K-means clustering and neural networks 7. The metrics to evaluate the performance of regressive and classification models on a testset: R2 as the explained variance, MAPE, accurcy, sensitivity and specificity in the confusion matrix, the ROC curve and AUC as cut-off point independent metric, and the model lift (curve) 8. The iterative apporach to build, govern, maintain and continuesly improve models.

Address

• Headquarters

  Vijzelstraat 721017 AmsterdamNoord-Holland, Netherlands