STUDYING DATA SCIENCE - University of Helsinki...Facultyof Science OUTLINE •Styles and levels of...
Transcript of STUDYING DATA SCIENCE - University of Helsinki...Facultyof Science OUTLINE •Styles and levels of...
Faculty of Science
STUDYINGDATA SCIENCE
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 1
Faculty of Science
OUTLINE
• Styles and levels of learning
• Study modes
• Teaching, learning and assessment methods
• Personal study plan in Data Science
• Model schedule of studies in Data Science
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 2
Faculty of Science 31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 3
STYLES AND LEVELSOF LEARNING
Faculty of Science
LEARNING STYLES
• People are different and have different ways of learning, i.e., different learning styles
• There are many models of learning styles, for example,o Learning styles based on which sense is used (auditory, visual,
kinaesthetic, tactile)o Learning styles based on different abilities (visual, aural, verbal,
physical, logical, social, solitary)
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 4
Faculty of Science
LEARNING STYLES (2)
• Learning styles based on different abilitieso Visual: prefer using pictures and imageso Aural: prefer using sound and musico Verbal: prefer using words (speech and writing)o Physical: prefer using the body, hands and sense of toucho Logical: prefer using logic, reasoning and systemso Social (interpersonal): prefer to learn in groups or with other
peopleo Solitary (intrapersonal): prefer to work alone and use self-study
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 5
Faculty of Science
LEARNING STYLES (3)
• By recognizing and understanding your own learning styles, you can use techniques better suited to youo Improves the speed and quality of your learning!
• People also have different approaches to (the level of) learningo Surface learningo Deep learning
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 6
Faculty of Science
LEVELS OF LEARNING
• Bloom's Taxonomy of Cognitive Levels of Learning (from 1958)o Knowledgeo Comprehensiono Applicationo Analysiso Evaluationo Synthesis
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 7
http://www.mrswatersenglish.com/2014/06/
Faculty of Science
LEVELS OF LEARNING REQUIRED
• Advanced learning skills• Manage your own studies
• High requirements• We do fail students!
• Deep level of learning• No grades with just remembering
• Learning in the focus• Very few obligatory events
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 8
Faculty of Science 31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 9
STUDY MODES
Faculty of Science
DIFFERENT STUDY MODES
• Lecture courseso Weekly lectures and exercise sessions + course examo May contain small reports, programs, etc
• Separate examso Reading for the exam and taking the examo Can be used as renewal or make-up exam of courses
• Project courseso Large project(s), typically team work
• Seminarso Report + presentation + review
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 10
Faculty of Science
COMPLETING COURSES
• Two wayso Lecture courseo Separate exam
• Both require advance registration
• Does not apply for projects and seminars
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 11
Faculty of Science
LECTURE COURSES
• Usually 5 credits• Consist of o Lectureso Weekly exerciseso Sometimes a project worko Typically one course exam; some courses have also a renewal
exam• Always check details from the course homepage!
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 12
Faculty of Science
SEPARATE EXAMS
• Formally independent of the lecture courses• Requirements are based on the material in the course
description• Duration 3.5 hours• Registration for the separate exams also in WebOodi• If you want the exam questions in English, contact the
examiner two weeks before the exam
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 13
Faculty of Science 31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 14
TEACHING, LEARNING, AND ASSESSMENT
METHODS
Faculty of Science
GENERAL ISSUES ON USED METHODS
• Teaching, learning and assessment methods used in courses at the departments vary a lot
• To know what kind of methods are used in a particular courseo attend the first lecture, ando check the course pages
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 15
Faculty of Science
LECTURE COURSES
• A basic rule: a 7-week course that yields 5 credits consists ofo 2*2 hours lectures per weeko 2 hours exercises per week
• Remember to schedule some 2 hours of independent work per every classroom or exercise hour (or even more)
• But there are many courses that are organised differently, e.g.,
o One lecture per week
o One introductory lecture + individual or group work
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 16
Faculty of Science
FORMS OF LECTURES
• Lecturer may have prepared lecture slideso Usually available on the course pages after the lecture (or even
before)o Students need to make notes only on the additional things
• Or lecturer just talks, gives maybe a few key words or writes something on the blackboardo Students need to make notes totally themselves
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 17
Faculty of Science
FORMS OF LECTURES (2)
• Lectures may also be interactiveo Discussion on some topics
§ With the whole group of students§ In small groups (3-4 students)§ Pairwise
o Ex tempore assignments§ Each student first tries to solve the task and then it is discussed
together
§ The solution is directly constructed together
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 18
Faculty of Science
FORMS OF EXERCISES
• Weekly home assignments/taskso Students try to solve the given tasks individually or together with
peer students before the exercise session
• In exercise sessionso One student is chosen to represent his/her solution to a tasko The tasks are discussed first in small groups and then each group
represents their solution to a tasko The solutions are discussed together, and the teacher may show
some example solutions to the tasks
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 19
Faculty of Science
FORMS OF SEMINARS
• Introductory session
• Schedules of presentations varyo One presentation per week during the whole term, reports ready
approximately a week before the oral presentationo Reports are written during the first period of the term and oral
presentations are held during the second period of the termo Reports are written during the whole term, and oral presentations
are held at the end of the term in one session
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 20
Faculty of Science
LEARNING METHODS AND STYLES
• Most people learno 10% of what they reado 20% of what they hearo 30% of what they seeo 50% of what they see and hearo 70% of what they talk over with otherso 80% of what they use and do in real lifeo 95% of what they teach someone else
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 21
(W. Glasser 1988)
Faculty of Science
APPROACHES TO LEARNING
• Being active while learning is better than being inactive
• Different approaches to learningo Surface learningo Deep learning
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 22
Faculty of Science
STUDY SKILLS
• In learning, we need different study skillso General study skillso Study skills that relate to learning particular contento Meta-cognitive learning skills (what to do in new
contexts/situations)
• Development of your own study and learning skills is important!
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 23
Faculty of Science
GRADING
• The grading scale is divided into six steps (0-5)• On courses, o to gain the lowest passing grade, 1/5, a student usually needs to
gain half of the maximum pointso 5/6 of the maximum points usually gives grade 5/5
• In MSc theses, a grading scale with 7 steps is used• Approbatur, lubenter approbatur (2/5)• Non sine laude approbatur, cum laude approbatur (3/5)• Magna cum laude approbatur (4 /5)• Eximia cum laude approbatur, laudatur (5/5)
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 24
Faculty of Science
ASSESSMENT IN COURSES
• Typically the maximum number of points in a lecture course is 60 pointso To get grade 1/5 you usually need to gather at least half of the
maximum points, e.g. 30/60 pointso To get grade 5/5 you typically need to gather 5/6 of the maximum
points, e.g. 50/60 points
• In some cases the maximum number of points can be higher or lower, too
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 25
Faculty of Science
ASSESSMENT IN COURSES (2)
• Different parts of the course can give you points for completing the courseo Typically:
§ at maximum 54 points from the exam,
§ at maximum 6 points from the exercises
o Can be:§ at maximum 30 points from the exam,
§ at maximum 30 points from the exercises
o Or something totally different!
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 26
Faculty of Science
ASSESSMENT IN COURSES (3)
• Different parts may have thresholds that have to be met in order to complete the courseo For example, you must get at least
§ half of the maximum points given the exercises,
§ half of the maximum points given in the exam, and
§ half of the maximum points given all together.
• Check always which rules apply for the course in question!
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 27
Faculty of Science
ASSESSMENT METHODS USED IN COURSES
• Examso Course, renewal or separate examo Home exam
• Essay, article• Computer-assisted exercises• Project work• Self-assessment• Peer-assessment
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 28
Faculty of Science
EXAMS
• Typically 3-5 questions per exam
• Length of the exams:o 2.5 h for a course examo 3.5 h for a separate examo 1-2 weeks for a home exam
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 29
Faculty of Science
EXAM QUESTIONS
• Different types of questionso Definitions of concepts and terms
§ Briefly, precisely: “What does the concept X mean?” or “What is the basic idea in ...?”
§ More thoroughly: “Explain in detail how the method X works?” or “Give a description of X and Y. What are their similarities and differences?”
o Giving examples§ “What is X? Give an example ...” or “Give an example on how the
algorithm works when ...”
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 30
Faculty of Science
EXAM QUESTIONS (2)
• Different types of questions (cont.)
o Applying some algorithms or methods in a given scenario, with given data§ “Let us assume that .... Explain what happens when ...”
§ “Give the values of X and Y, when the input to algorithm A is Z.”
o Analysis or evaluation of a method
§ “In which cases method X works well?” or “How system X can avoid possible problems?”
o Comparison of two or more methods
§ “When is algorithm X faster than algorithm Y?”
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 31
Faculty of Science
ASSESSMENT IN SEMINARS
• Based on o the report written on the given topic,
o the oral presentation and
o the activity of the participation in held sessions and discussions
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 32
Faculty of Science 31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 33
PERSONAL STUDY PLAN IN DATA SCIENCE
Faculty of Science
PLANNING OF STUDIES
• The size of the MSc degree is 120 credits• The optimal study time is 2 yearso 2 years = 4 termso About 120/4 = 30 credits per term
• A course that yields 5 credits requires, for example, a minimum of about 125 hours of work
• Advice in making a PSP in Data Science: o Education Coordinator Reijo Sivèn, [email protected], D239
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 34
Faculty of Science
PLANNING OF STUDIES (2)
• Only few compulsory courses (35 credits)
• Rest of the courses you need to select
o to fulfil the degree requirements
o to support your thesis
o to be interesting for yourself
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 35
Faculty of Science
PLANNING OF STUDIES (3)
• The basic planning rule:
o Schedule some 2 hours of independent work per every classroom or exercise hour
• The number of lectures and other contact teaching per course (and credit) varies
o Uncommonly few teaching hours in relation to the number of credits a course yields => the portion of independent work is even larger than described above
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 36
Faculty of Science
PLANNING OF STUDIES (4)
• Why?
o Unique experts at graduation§ Different course selections
§ Own, unique thesis
o Allows unexpected combinations
o Students learn to manage their own plans and schedules
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 37
Faculty of Science
IMPORTANT GUIDES AND LINKS
• Instructions for students: http://guide.student.helsinki.fi/en• My studies: https://student.helsinki.fi/info/login
• Weboodi: https://weboodi.helsinki.fi• Course pages: https://courses.helsinki.fi
• Web pages of the programmeo https://www.helsinki.fi/en/programmes/master/data-science
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 38
Faculty of Science 31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 39
MODEL SCHEDULE OF STUDIES
IN DATA SCIENCE
Faculty of Science
TEACHING PERIODS 2017-2018
• 4 periods (7 weeks teaching + one exam week)o I Period: September 4 - October 22 (exams: 23.10.-29.10.)o II Period: October 30 - December 17 (exams: 18.12.-22.12.)
o [Intensive period in January (2.1.-14.1.)]o III Period: January 15 - March 4 (exams: 5.3.-11.3.)o IV Period: March 12 - May 6 (exams: 7.5.-13.5.)o Intensive period in May (7.5. – 31.5.)
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 40
Faculty of Science
PERIOD 1 IN 2017-2018
• Core courses:o Introduction to Data Scienceo Orientation to Data Science Studieso Academic Writing 1 (continues on Period 2; groups 9 and 10)
• Elective courses:o Computational Statistics Io Design and Analysis of Algorithmso Introduction to Artificial Intelligence
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 41
Faculty of Science
PERIOD 2 IN 2017-2018
• Core courses:o Bayesian Inferenceo Distributed Data Infrastructureso Introduction to Machine Learningo Academic Writing 1 (continues from Period 1; groups 9 and 10)
• Elective courses:o Computational Creativityo Data Compression Techniqueso Deep Learningo High Dimensional Statisticso Introduction to Big Data Management
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 42
Faculty of Science
PERIOD 3 IN 2017-2018
• Core courses:o Data Science Project I (continues on Period 4)o Data Science Seminar I (continues on Period 4)o Academic Writing 2 (for Centre Campus programmes, continues
on Period 4)
• Elective courses:o Cloud and Edge Computing
• Seminars:o Big Data Management, Experimental Algorithms, Applied Discrete
Algorithms, Deep Learning for Natural Language Processing
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 43
Faculty of Science
PERIOD 4 IN 2017-2018
• Core courses:o Data Science Project I (continues from Period 3)o Data Science Seminar I (continues from Period 3)o Academics Writing 2 (for Kumpula and Viikki Campus
programmes)o Academic Writing 2 (for Centre Campus programmes, continues
from Period 3)
• Elective courses:o Interactive Data Visualizationo Advanced Course in Bayesian Statisticso Advanced Course in Machine Learning
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 44
Faculty of Science
REGISTRATION FOR COURSES AND EXAMS
• Registration is done in Weboodi
• Registration for lecture courses typically start 2-4 weeks before the beginning of the period
• Registration for separate exams must be done at latest 10 days before the examo If you want the exam questions in English, contact the examiner
two weeks before the exam
31.8.2017Orientation of the Master’s Programme in Data Science / Pirjo Moen 45