Decision tree methods

Decision tree methods (TRM)

This cour­se gives a deep insight into the metho­do­lo­gy of decis­i­on trees. Decis­i­on trees are robust, fast and under­stan­da­ble and the­r­e­fo­re popu­lar for data explo­ra­ti­on and fore­cas­ting with lar­ge and hete­ro­ge­neous data sets.

Con­tents

• Basics of decis­i­on trees: divi­si­on rules, pru­ning, ana­ly­sis
• CART
• CHAID
• Ran­dom Forest
• Boos­ted trees
• Over­view of other methods: ID3, FACT, QUEST

Type
Our method trai­ning cour­ses are desi­gned to be soft­ware-inde­pen­dent and cross-indus­try. We attach gre­at importance to a lively and inspi­ring atmo­sphe­re in the trai­ning and the­r­e­fo­re pre­sent examp­les from various appli­ca­ti­on are­as; soft­ware may also be used.

The cour­se is open to dis­cus­sion and con­tri­bu­ti­ons from par­ti­ci­pan­ts. Our trai­ners respond to all ques­ti­ons from par­ti­ci­pan­ts and ans­wer them at appro­pria­te points during the cour­se.

Par­ti­ci­pan­ts will recei­ve all trai­ning docu­ments in PDF for­mat after­wards.

Key­words
Bag­ging, boos­ted trees, boos­ting, C&RT, CART, CHAID, entro­py, explo­ra­ti­on, miss­ing values, miss­ing value repla­ce­ment, GINI, infor­ma­ti­on gain, pre­dic­tion, ran­dom forest, sto­cha­stic gra­di­ent boos­ting, divi­si­on rules, varia­ble weight­ing, ID3, FACT, QUEST

Pre­re­qui­si­tes
This cour­se assu­mes a basic know­ledge of data mining.