zhongjiu

1. Statement The motivation of this blog is mainly driven by a number of LinkedIn connections requesting information about this HOT programme, and also my personal mark on my academic journey. The objective of this article is to provide an insightful, personal reflection on this MSc programme. Statement: the following article only represents the authors’ PERSONAL opinion ONLY. 2. Background of the MSc Business Analytics Programme This programme started in academic year 2015/2016. according to unofficial statistics (word of mouth from alumnus), there were 40 students, 60 students and 81 students in year 2015/2016, 2016/2017, and 2017/2018 respectively. The initiative for Imperial College Business School to establish this programme is to meet the demand from the partnership industry professionals who want to upgrade their analytics skills. This year 2018/2019 update: There are around 80 full-time students on campus, and approximately 40 students enrolled part-time online-delivery course (NEW programme from year 2018/2019) 3. Why do I choose to study this programme? A litter introduction about myself, I did my undergraduate degree in Manchester, UK, studied BSc Economics, specialised in Econometrics and Mathematical Economics during my final year. During my second-year study of Econometrics, I had a science fiction taste of statistical modelling and the state of the art approach in time series forecasting prediction. After that, I interned at a research unit doing health econometrics project. This internship opportunity exposed me to real data analytics projects and data science overall. The post effect for me was to select more quantitative modules in my final year of study and decided to pursue a postgraduate degree that enables me to step into the era of Data Science. After some research, Imperial College MSc Business Analytics programme was (and still is) the best programme I can find in the UK (I was not considering other countries due to personal reasons). Furthermore, I was not thinking to pursue an academic career. MSc Business Analytics programme seems to be the perfect combination of technique and business application, opening doors in both technology and traditional business opportunities for me. (Data Science / Big Data is a hot topic, this also interests me too) 4. Key take away • Gained an insightful industry understanding on how data science or data analytics is applied across different industries • A great “saddle” point to switch between analytics consulting and technology industry (The major destination for this programme so far, the data is only updated until now) 5. Programme structure 5.1 Content structure Here is the official link to the programme content structure, it is worth a read: https://www.imperial.ac.uk/business-school/programmes/msc-business-analytics/programme/modules/ 5.1.1 July - August (Pre-Session) You have to take typical business school training courses, which consist of Accounting, Finance, Mathematics. All the modules are delivered online - The Hub, Imperial internal studying platform. The assessment format is MCQ, and take place at the end of September. The final result is PASS or Fail. 5.1.2 September Term There are only two INTENSIVE modules, Maths and Statistics foundations (MSF) AND Data Structures and Algorithms (DSA). • MSF refreshes your basic mathematics and statistics, it is more or less like 1st to 2nd year university level. The key for this course is to learn how to use Github, version control, Rmarkdown, R language and many data science R libraries (i.e. ggplot2, dplyr and etc). We were given access to Datacamp and required to obtain certain XP scores as part of the assessment. • Highlight: the final 50% of the assessment is to write R code and test your statistic understanding • DSA provides you sufficient training on the computer language programming. The main programming concept covers from the basic loop, recursion, to OOP. In addition, we also covered some graph theory, and algorithm complexity. Python programming language is further introduced through the entire programme. • Highlight: this course content goes super-fast and does not really go deep into each section. 50% of the assessment is computer-based python programming test. Summary: This month is probably one of the toughest periods because of the fast pace of new content being covered. Python and R were introduced. Highly recommended to immerse yourself in studying. 5.1.3 Autumn Term (Oct - Dec) All the modules are mandatory, so you will see your classmates everyday with the identical timetable. Optimisation and Decision Models • Linear Programming, Convex optimisation, Quadratic programming, Duality, Constraint optimisation and Algorithm • Planning different resources in different departments using optimisation models • Assessment: 50% coursework using the toolkit, 50% paper-based exam • Toolkit: Excel (solver), AMPL • Highlight: Enthusiastic professor, Wolfram Wiesemann Statistics & Econometrics • Regression analysis, Hypothesis testing, Probit/Logit regression, Panel data analysis, and Instrumental variable methods • Build a simple supervised machine learning model to do prediction • Assessment: 50% coursework using the toolkit, 50% paper-based exam • Toolkit: R • Highlight: The pace is super-fast! Fundamentals of Database Technologies • Relational database, database normalisation, parallel computing concept, query optimisation • Assessment: 50% coursework using the toolkit, 50% computer-based exam • Toolkit: postgreSQL, Pgadmin, Spark on Databrick platform • Highlight: It is more about independent study, if nothing has been changed Machine Learning • Classic machine learning algorithms are introduced, KNN, tree-based algorithms, hierarchical clustering, k-means clustering, decision tree. Important machine learning concepts: variance bias trade-off, cross-validation, boosting algorithm • Assessment: 50% coursework using the toolkit, 50% paper + computer based exam • Toolkit: R and Python • Highlight: Enthusiastic professor, Wolfram Wiesemann again! Network Analytics • Graph theory, matching, Diffusion models, influence in Social networks and Community detection • Assessment: 50% coursework using the toolkit, 50% paper + computer-based exam • Toolkit: Python (package: networkx) • Highlight: Broad coverage of topics, time consuming and difficult coursework Summary of the term • This term is the foundation for all the following courses, it is better to get handy with toolkits sooner than later. • Every week consists of 2 to 3 coursework to handle in on next Monday. • All of them have exams after the New Year, and some of them are also computer-based exams. 5.1.4 Spring Term (Jan - April) This term consists of compulsory modules and electives The two compulsory modules are: • Analytics in Business • Visualisation Analytics in Business • Consultancy concept in business and data analysis business case study • Assessment: class quiz and projects • Toolkit: Free to use any (we tend to use R more than Python) • Highlight: business case study, VERY tricky quiz Visualisation • Design theory, Colour theory, visualisation case studies • Assessment: 50% coursework, 50% paper-based exam • Toolkit: Tableau • Highlight: Two different lecturers for this module My selected modules: • Advanced Machine Learning • Logistics & Supply Chain Analytics Advanced Machine Learning • More machine learning algorithms: LDA, QDA, Lasso & Ridge regression, Principal Components, Support vector machine and other advanced machine learning algorithms • Assessment: 50% coursework using toolkit, 50% computer + paper-based exam • Toolkit: Free to use either R or Python • Highlight: Practical coursework, very statistical based exams, and a great professor Logistics & Supply Chain Analytics • Time series forecasting, logistics inventory planning • Assessment: coursework, presentations and inventory simulation game • Toolkit: R • Highlight: The simulation game is somehow a love-hate relationship? definitely a highlight 5.1.5 Summer Term Block 1 (April-May) All the modules are selective modules and I selected the following: • Digital Marketing Analytics • Data Management and Ethics • Workforce Analytics Digital Marketing Analytics • Database marketing, CRM, customer churn case, web traffic analytics, the recommendation system • Assessment: 100% coursework • Toolkit: Google Analytics, postgreSQL, R or Python • Highlight: Extremely practical case study and coursework. Data Management and Ethics • ETL process for data, Big Data infrastructure, MapReduce programming • Assessment: Two 50% coursework • Toolkit: Pyspark and NoSQL(MongoDB) both on Microsoft Azure • Highlight: A good understanding and practical experience with ‘big data’, cloud computing. Workforce Analytics • Natural language process application, network/clustering analytics, hands-on experience in feature engineering, and final predictive model implementation • Assessment: 30% group project, 70% individual project • Toolkit: Python • Highlight: I believe it is the most applied module through the programme. You will not learn much technique content, but gain a comprehensive practical experience. 5.1.6 Summer Term Block 2 (June-Sep) This term has overall three different combinations: 1. Capstone project + Business Analytics report 2. Work placement + Business Analytics report 3. Individual research report This year, we have around 50% in combination 1, 45% in combination 2, and only 5 % in combination 3. The setting is the following: you need to secure a work placement (Internship) before May, you will have options amongst the combinations. Otherwise, you choose 1 or 3. Individual research report is a more research-based project, it is designed for students who are interested in academic or a particular topic. I went for option 2, the simple reason is to obtain some financial reward supporting my travelling plan. 5.2 Cost of the programme • Tuition fee: £27000 • Accommodation fee: £10000 - £12000 • Living expenses: £500 * 12 = £6000 • Other entertainment expenses: £3000 - £5000 I personally believe the expense range from £50,000 to £60,000 ON AVERAGE 6. Cohort background 6.1 Geographical background I categorise in the following regions: Asia - 42 • China 27 • India 5 • Singapore 4 • South Korea 3 • Thailand 2 • Malesia 1 Europe 28 • The major group is Germany USA - 2 South America - 4 Middle East -2 Canada – 1 6.2 Academic background • 40%-50%: Engineering & computer science background • 20%-30%: Economics background • 20%-30%: Business management related background 6.3 Work experience background • Most Asian students’ work experience ranges from 0 to 5 years, highly right skewed distribution • Other groups have on average 2 to 3 years more work experience than Asian students • Highlights: there are several industry professionals in this programme. They worked in Banking, Marketing and other industries for more than 5 years, they come here to either bring new blood to their organisation or seek to change careers. 7. Career opportunities The business school provides a WEALTHY of resources to support you through your time at Imperial. Here are the reasons: • Many career events through the academic year, especially frequent between Sep and Dec • Unlimited appointments with career consultant at both Business School (Business School students ONLY) and Imperial College overall career service (for all imperial students) • Appointments cover from CV to one-to-one mock interviews, it can be both online (face-to-face at the Imperial College office) and off-line (Mainly Skype) • Uncountable company's presentation (Big banks, consultancy, technology companies) Some highlights of BA programme. • Many consultancy firms come to recruit this programme students (Just to name well-known companies: OC&C strategy, KPMG (Forensic Data Analytics), Monitor Deloitte, BCG Gamma) There are several Data Challenge or Datathon events through the year, some of them I remembered • Data Science Society Data Challenge • OC&C strathon at Oct Personal recommendation: Try to attend these data challenge activities because of the following reasons • A great experience in data science application across different industries • Enhance your data wrangling skills, either R or python. • Improves your employability in recruitment assessment There are more companies starting using analytics case study to assess a candidate. The process is the following: • You are given a dataset and a business problem to solve • Submit your code and presentation within one or two weeks • If you are selected, present your insight from data to a number of executives If you pass the analytics challenge, you will be interviewed 1 to 2 more rounds in general. Therefore, I highly recommend participating in these data challenges. 7.1 Geographical destinations of this year cohort I categorise two streams of destinations and possible reasons: Go back to their motherland • Workfare is somehow better than in the UK • Having boy/girl friends in motherland • Do not enjoy the environment in London (i.e. weather, food and etc) Stay in the UK • Value the overseas work experience • Work life balance is better in the UK than motherland • Enjoy the culture and environment in London 7.2 Informal destination of this year cohort Overall, consultancy and technology are the major destinations Here is a list of companies having at least two students from our programme: • OC&C strategy • YIMian Data • Concentra Analytics • Expedia Group Other offers I know: • Amazon • Vodafone • Bain & Company • Daimler AG • Accenture • Context • Prudential • Satalia • Aviva There are quite a few who will work in start-ups in London, which I do not really remember the names of the start-ups. the detail report will be available at Imperial website: https://www.imperial.ac.uk/business-school/programmes/msc-business-analytics/careers/employment-report/ 7.3 My career options I obtained offers in Banking, Insurance, Fintech and consulting, the destinations range from London, Madrid, Beijing and HK. I am about to work for a UK insurance company as a Data Scientist. 8. School social experience The programme team organises several social events for students, drink & food are included for free Here are the social activities I can remember now: Before Christmas • Alumnus meet up at university union bar (BA students) • Welcome reception (ALL Business School students) at the National History Museum • Scape Room team building social (BA students) • End of term Christmas party at Central London (ALL business school students) • Christmas winter wonderland at Hyde Park London (BA students) After New Year • Clubhouse for free bowling, food and Karaoke at Feb (BA students) • End of the year summer ball in a hotel (ALL business school students) 9. My life experience at this programme Since life experience varies so much depending on individuals The highlight of my experience in London this year • Christmas trip to Finland, first time to arctic circle under 20 Celsius degree • Easter Break to Turkey for a road trip - drove more than 3000km • Watched around 10 musicals in London. i.e. Hamilton, Les Misérables and etc • Cycling trip in Brighton for an entire day • Hiking at Peak District • Many visits to London landmarks for lunch and dinner I personally really enjoy the vibrant atmosphere and diversity in London. Most important thing for me is I made some lifelong friends in this programme. 10. Questions & Answers Here are some of the questions from Adit Kulkarni, the student at year 2018/2019. Thank you for his contribution! Q: How technical is the course overall? A: The programme covers a very board range of subjects, which are listed quite explicitly above. Theoretically, every module has shown students the theories, conclusion and relevant references. However, it does not focus on theory. For example, OLS regression model does not require students to derive the conclusion or remember the assumptions. Instead, it taught you how to use these techniques in a real-world problem and the interpretation of the regression coefficients. Practically, every module heavily consists of coursework, which trains you well in applying the theoretical knowledge. Every Core module has 3 to 5 coursework on average, such amount of workload does put you in a good position of technical. My conclusion: YES, it is a very technical programme if you are keen on applied science. Q: Does this programme provide suitable preparation to pursue PhD/another MSc in Machine Learning? A: This is a follow-up question from above. It does not provide you with a solid theoretical training, which you can obtain from a more academic programme. However, whether pursuing a Ph.D. is your personal determination or career aspiration, because you can of course spend your time to read all the references and study it further. Moreover, Business School and other schools do have many fully-funded Ph.D. programme advertisement across the academic year, it is entirely up to you what you would like to achieve. From my personal experience, Machine Learning professor Wolfram, Wiesemann advertised a full-funded Ph.D. programme during his lecture, Autumn term. I had a few conversations with him along with other interested students from BA programme. Here are some checkboxes to determine whether you should do a Ph.D. Importance is ranked in a descending order. 1. You want to have an academic career 2. You MUST be interested in your chosen topics (i.e. Machine Learning) 3. There is a professor at the School who would like to supervise you (In general, you and the professor are both interested in the chosen field/topic) My conclusion: Not really, but you can take your chance to work this out. Q: What career options do you have? A: In general, it opens career paths in both consultancy industry and technology industry. The typical career titles can be Data analyst, Analytics consultant, Business Analytics, and Data Scientist. A more detailed information is available at Imperial website: https://www.imperial.ac.uk/business-school/programmes/msc-business-analytics/careers/employment-report/ Q: Do you have much time to go out? A: This significantly depends on your priorities or goals at this programme and your time management skill. In general, it is an intense programme. Therefore, you would not have much time to go out, compare to undergraduate. 11. Recommendations The question I have in mind is: Q: If I could restart this programme, what would I have changed? A: I would immerse myself in studying at least until January (Spring Term) rather than head into job hunting from the beginning. I had missed out some good opportunities due to the lack of technical expertise, and the earlier failure stopped me to apply for these positions again within the probation period (In general 6 to 12 months) 12. Additional resources This is a review from the student at first batch, Jonathan Zimmermann: https://www.quora.com/How-is-the-MSc-in-Business-Analytics-program-at-Imperial-college-London