IFTA2020: How to Build Investment Strategies with AI and Machine Learning

Recent advan­ces in AI and machi­ne lear­ning are fin­ding prac­ti­cal appli­ca­ti­ons across many indus­tries, not least the finan­ce indus­try. In par­ti­cu­lar, rese­arch into the use of AI and machi­ne lear­ning as a new tool for stock invest­ment has been very active.

 I’ll show how AI and machi­ne lear­ning can be appli­ed to the pro­blem of dis­co­ve­ring and imple­men­ting stock invest­ment strategies.

As an exam­p­le, I’ll intro­du­ce an exam­p­le of deve­lo­ping machi­ne lear­ning based fore­cas­ting methods while focu­sing on the per­spec­ti­ve  of pri­ce fluc­tua­ti­on pat­terns, which is the phi­lo­so­phy of tech­ni­cal analysis.

Dr. Naka­ga­wa is a quan­ti­ta­ti­ve ana­lyst in Inno­va­ti­on Lab at Nomu­ra Asset Manage­ment Co Ltd. from 2018.
He enga­ges in the R&D of cut­ting-edge tech­no­lo­gies such as AI, machi­ne lear­ning, and Fin­Tech and their appli­ca­ti­on to the asset manage­ment busi­ness.
He recei­ved a B.A., in Eco­no­mics from Kyo­to Uni­ver­si­ty in 2012, an MBA from the Uni­ver­si­ty of Tsu­ku­ba in 2015, and a Ph.D. in Busi­ness Admi­nis­tra­ti­on in 2020 from the Uni­ver­si­ty of Tsu­ku­ba.
From 2012, He was a risk mana­ger and quan­ti­ta­ti­ve invest­ment stra­tegy fund mana­ger.
He mana­ged quan­ti­ta­ti­ve funds inves­t­ing in Japa­ne­se Long-Short Equi­ty, Glo­bal Equi­ty, and REIT. He deve­lo­ped a lot of quan­ti­ta­ti­ve equi­ty and mul­ti-asset allo­ca­ti­on models.
Not only does he have work expe­ri­ence as a quant, but he also has many aca­de­mic pre­sen­ta­ti­ons and papers on arti­fi­ci­al intel­li­gence, machi­ne lear­ning, and finan­cial engi­nee­ring inclu­ding top-tier AI aca­de­mic con­fe­ren­ces.
His rese­arch inte­rest lies in impro­ving the method of finan­ce by machi­ne lear­ning and, on the con­tra­ry, deve­lo­ping machi­ne lear­ning by the method of finan­ce.
He tea­ches at Tokyo Insti­tu­te of Tech­no­lo­gy, the Uni­ver­si­ty of Tsu­ku­ba, and Wase­da University.

Video nur für eingeloggte/angemeldete Benutzer