42184 Data science inden for mobilitet

2019/2020

Kursusinformation
Data Science for Mobility
Engelsk
5
Kandidat
E1A (man 8-12)
Campus Lyngby
Forelæsninger, gruppearbejder og online undervisning (for eksempel iPython notebook)
13-uger
Bedømmelse af opgave(r)/rapport(er)
7-trins skala , intern bedømmelse
42577
42577 Introduction to Business Analytics
02402/02403/02105
Francisco Camara Pereira , Lyngby Campus, Bygning 116, Tlf. (+45) 4525 1496 , camara@dtu.dk
Filipe Rodrigues , Lyngby Campus, Bygning 116, Tlf. (+45) 4525 6530 , rodr@dtu.dk
Stanislav Borysov , stabo@dtu.dk
42 Institut for Teknologi, Ledelse og Økonomi
I studieplanlæggeren
Dette kursus giver den studerende en mulighed for at lave eller forberede et projekt som kan deltage i DTUs studenterkonference om bæredygtighed, klimateknologi og miljø (GRØN DYST). Se mere på http://www.groendyst.dtu.dk
Overordnede kursusmål
This course introduces the portfolio of data science tasks and techniques necessary for exploring, manipulating, visualizing and analysing data (descriptive analytics), as well as for building prediction models using machine learning (predictive analytics) that can be used to gain insights and support decisions (prescriptive analytics). It is designed with Management Engineering students in mind (i.e. some programming background, but not the core skill), particularly – but not exclusively - those related to studies on mobility and logistics and business analytics. Therefore, it contains a strong hands-on component, with specific real-world cases from mobility and business contexts.

The course also includes an introduction to Python programming, data wrangling, problem formulation, and the basic suite of machine learning algorithms.
Læringsmål
En studerende, der fuldt ud har opfyldt kursets mål, vil kunne:
  • Run Python scripts that load and analyse small/medium-sized datasets
  • Convert a raw dataset into an actionable form to solve a concrete problem
  • Apply basic data structures and algorithms to manipulate data
  • Relate available problems and data in a mobility context with techniques to tackle them
  • Extract and analyse insights from the application of methods for descriptive and predictive analytics
  • Visualize complex temporal and spatial patterns
  • Argue for the choice of appropriate data analysis and predictive analytics algorithms
  • Appropriately train and test a model to answer a problem
  • Critically evaluate the results of a data sciences analysis and recommend actions
  • Explain important data mining concepts, such as overfitting, bias, regularisation, etc.
Kursusindhold
The classes are taught in an interactive manner, with theoretical parts, intermingled with practice with Jupyter Notebooks. 

Main topics are: Introduction to Python programming and Pandas, data visualisation, forecasting and regression models, classification, clustering, dimensionality reduction. The methods will be exemplified through different cases within e.g. transportation, business management and marketing.
Litteraturhenvisninger
Udrag fra:
"Data Science from Scratch", Joel Grus
"Python for Data Analysis", Wes McKinney
"The elements of statistical learning", Trevor Hastie, Robert Tibshirani and Jerome H. Friedman
Sidst opdateret
09. april, 2019