Bonn Data Science Meetup: Forecasting and Vehicle Routing Optimization

    Wann:
    01.02.2018 um 19:00 – 22:00
    2018-02-01T19:00:00+01:00
    2018-02-01T22:00:00+01:00
    Wo:
    Post Tower
    Charles-de-Gaulle-Straße 20
    53113 Bonn
    Deutschland
    Preis:
    Kostenlos

    Nach einer kleinen Pause starten wir im neuen Jahr mit zwei Vorträgen des zentralen Data Analytics Teams der Deutschen Post/DHL. Anja und ihre Kollegen stellen zwei Themen aus ihrer Praxis vor: Vorhersage von Zeitreihen und Routenoptimierung. Die Vorträge werden auf englisch gehalten, ihr findet unten eine kurze Vorschau, was euch erwartet. Im Anschluss wird wie immer Zeit für Diskussionen und Networking sein und auch für das leibliche Wohl ist gesorgt.
    Wir freuen uns auf rege Beteiligung und wünschen euch allen bis dahin schöne und erholsame Feiertage und einen guten Start ins neue Jahr. Forecasting @ Deutsche Post DHL
    Forecasting is a very important topic at Deutsche Post DHL Group. It helps the business across a wide range of tasks such as work force and network planning, capacity management, mid-term procurement and controlling in finance. We will first give a short introduction to time-series forecasting tailored to people with analytical background but limited time-series experience. This part will cover our experience on best practices when working with time-series as well as some insights into the methods we usually use. Afterwards, we will go deeper into the methodological part of our projects and present some challenges we have faced during the development of our projects. In this part, we will focus on tricks and ideas that have helped us to increase accuracy of our forecasts. Finally, we discuss some newer developments in the field that we plan to test in our next projects.

    Vehicle Routing Optimization @ Deutsche Post DHL
    On top of being closely intermingled with various machine learning tasks, optimization theory provides the technology and tools to solve data-driven decision making problems in its own rights. In this talk, we first provide an introduction to Continuous and Integer Linear Optimization and how to implement these with Python. We discuss some application areas where mathematical optimization is used for different sectors and then outline a class of Vehicle Routing Optimization problems from ongoing projects and will discuss the challenges we face at Deutsche Post DHL Group and the approaches we take for the solution.