week 3 not good

• Big data, Data analytics and Machine learning
• Data


  What is data? Why do we need data?
• You might remember the qualitative data from our lectures on "Descriptive statistics" an "Regression" in the Lent term
• Data, descriptive statistics and inferential statistics
• Significance of data in Accounting and Finance
• Technically speaking, we are talking...
• Big Data


  A term used to describe the phenomenon of creating, curating and exploiting extremely large data sets to reveal patterns, trends and associations
• Many of these big data sets are so large that they cannot easily be held physically in one location and so they are stored in the Cloud or in multiple locations on large servers (and research is currently underway to see if these can be stored in the oceans)
• Big Data can be used to analyse the behaviour of groups and individuals(!) or to spot trends or associations that could be useful to society (such as health-related issues)
• Where does the data come from?

  • Mobile phones
  • Internet interactions, including search engines (Google)
  • Social media (FB, Instagram)
  • GPS devices (Google maps, Tomtom)
  • Smart speakers (Alexa, Google spot)
  • Buying behaviours (Amazon)
• Notice that there are some practical problems; Storage, Processing capacity, Timeliness of analysis
• e.g. One random Spreadsheet we worked with in the Lent term
• Everywhere! Are you being tracked?
• Characteristics of Big data and their application in Finance and Accounting

  • Velocity: speed of data generation and processing
  • Volume: amount generated / processed
  • Variety: structured vs. unstructured (free text) data
  • Value: monetisation and other uses
  • Veracity: trustworthiness (e.g. biases, abnormalities, and decision-making
  • Variability: changes in data
  • Visualisation: summarising
• Applications of Big data in FinTech in details will be covered in the last session
• Data analytics


  Allows us to analyse raw data and extract useful trends / insights from the data which helps individuals and firms to make decisions
• The type of data analytics depends on the type of information required to make a particular decision
• Remember the lab sessions and the group project from the Lent term?
• The analysts task is to comment on the volatility of stocks, not to make an investment decision. Irrespective of the findings, a client might still decide to invest in the most volatile stock because of the expectation that it would yield the highest return based on the findings of some other analyst using a different data set
• Analytics and research questions
• What kind of data analytics can you use in the following situations?

  1. A leading firm is in the headlines due to a scandal. You are considering shorting the stocks of the firm

  2. You are the Consultant for Strega's & Strano's chain of restaurants. Your task is to find out Why has the revenue dropped in the last 6 months? What can be done to increase the revenue?

  3. You are the CEO of Beer Jolly Good which is particularly popular among consumers with low disposable income. You are considering increasing the price by 50%, but you are afraid that sales will drop substantially
• Types of data analytics

  • What has happened? Descriptive analytics
  • Why did it happen? Diagnostic analytics
  • What will happen if....? Predictive analytics
  • What to do to make it (not) happen? Prescriptive analytics
• Predictive data analytics


  The art of building and using models to make predictions based on patterns extracted from historical data
• Some applications of predictive data analytics

  • Price prediction
  • Diagnosis and dosage prediction
  • Risk assessment
  • Propensity modelling
  • Document classification
• The patterns are extracted through automated processes called machine learning
• Machine learning (ML)


  The automated process for extracting patterns from data
• Supervised Machine Learning (SML)

  Used to build models used in predictive data analytics applications. SML techniques automatically learn a model of the relationship between a set of descriptive and target feature, based on historical examples
• More the number of historical examples, better will be the learning
• Now, you know where "Big data" comes in
• Steps in SML


  1. Learning

  2. Predicting
• Other types of MLs include unsupervised learning, semi-supervised learning and reinforced learning
• How does ML work?

  ML algorithms search through possible prediction models and searches for the model that explains the relationship between the descriptive and a target feature in a dataset. The criteria for search is to look for consistency
• This example gives a brief overview of its modus operandi and also the application of ML in Accounting and Finance
• For further details read Kelleher et al. (2020), Chapter 1
• Unsupervised ML (UML)


  We use UML when we do not have a target feature. Therefore, we model the underlying structure within the descriptive features in a dataset
• We can look at it as a way of feature generation
• Reinforced ML (RML)


  RML is used to control the behaviours of autonomous systems. e.g. training robots
• As the name suggests, we are 'reinforcing' the learning
• Brainteaser


  1. W&W's has launched a new job portal. This portal has become very popular particularly among students. The advertisers have started putting ads on the portal due to the high traffic and on-screen time. Discuss the following with your teams

  2. What kind of adverts will the advertisers consider putting on the portal?

  3. What kind of data will it generate?

  4. What can W&W's find out from the data? What kind of data analytics can they use?

  5. What kind of MLs can be used in the given context? What can be the potential target? What if we do not have a target?

  6. Be creative! We can literally go to town with this! You have 10 mins to discuss with your peers
• Strategy Consulting firms, data sales and decision-making