Fintech and Big Data5 min
Fintech refers to technology applied to financial services. A major driver is Big Data, defined by Volume (size), Velocity (speed/latency), and Variety (structure). Data comes from traditional sources and alternative ones like individuals (social media), business processes (corporate exhaust), and sensors (Internet of Things). Data Science involves processing this data through capture, curation, storage, search, and transfer, and visualizing it using tools like word clouds for unstructured inputs.

Key Points

  • Fintech applies technology to financial services.
  • Big Data sources: Traditional, Social Media, Corporate Exhaust, Internet of Things.
  • Characteristics: Volume (scale), Velocity (latency), Variety (structure).
  • Data Science processes: Capture, Curation, Storage, Search, Transfer.
Artificial Intelligence and Machine Learning5 min
AI simulates human cognition. Neural networks process information similarly to the human brain. Machine Learning (ML) allows computers to learn from data without explicit programming. Supervised learning uses labeled training data to model outputs, while unsupervised learning finds structure in unlabeled data. Deep learning uses multi-layered neural networks. Models must avoid overfitting (fitting noise) and underfitting (missing patterns).

Key Points

  • AI simulates human cognition; Neural Networks mimic brain processing.
  • Machine Learning: Algorithms learn from data (Training, Validation, Test sets).
  • Supervised Learning: Labeled inputs/outputs.
  • Unsupervised Learning: Unlabeled inputs, finds structure.
  • Overfitting: Model is too complex, fits noise.
  • Underfitting: Model is too simple, misses patterns.
Fintech Applications in Investment Management5 min
Fintech applications include Text Analytics (analyzing word frequency), Natural Language Processing (interpreting human language/speech), and Risk Analysis (stress testing with real-time data). Algorithmic trading automates execution based on rules, including High-Frequency Trading (HFT). Robo-advisors provide automated, low-cost investment advice, typically using passive strategies, though they may struggle to explain recommendations during crises.

Key Points

  • Text Analytics: Analyzes unstructured text/voice.
  • NLP: Interprets human language (e.g., speech recognition).
  • Algorithmic Trading: Automated execution; HFT exploits intraday mispricing.
  • Robo-advisors: Low cost, passive/conservative portfolios, automated risk profiling.
Distributed Ledger Technology5 min
DLT creates shared databases with consensus mechanisms. Blockchain is a DLT recording transactions in sequential blocks secured by cryptography and miners. Networks can be permissionless (open, trustless) or permissioned (restricted). Applications include Cryptocurrencies (electronic medium of exchange), ICOs (capital raising), Smart Contracts (self-executing), Tokenization (electronic proof of ownership), and clearing/settlement efficiency.

Key Points

  • Distributed Ledger: Shared database with consensus mechanism.
  • Blockchain: Sequential blocks, miners verify transactions.
  • Permissionless vs. Permissioned networks.
  • Applications: Cryptocurrencies, ICOs, Smart Contracts, Tokenization.
  • Benefits: Faster clearing/settlement, reduced counterparty risk.

Questions

Question 1

Which of the following best defines 'Fintech'?

View answer and explanation
Question 2

Data generated by business processes such as bank records and retail scanner data is best described as which type of alternative data?

View answer and explanation
Question 3

Which of the following characteristics of Big Data refers to the speed at which data is communicated?

View answer and explanation
Question 4

If a dataset has a size of 5,000 terabytes, this is equivalent to:

View answer and explanation
Question 5

Real-time stock market price feeds are characterized as having:

View answer and explanation
Question 6

Which of the following is considered an unstructured form of data?

View answer and explanation
Question 7

In the context of data science processing methods, 'Curation' refers to:

View answer and explanation
Question 8

Which visualization technique is most appropriate for illustrating the frequency that specific words appear in a text sample?

View answer and explanation
Question 9

Computer systems that are programmed to simulate human cognition are best described as:

View answer and explanation
Question 10

In Machine Learning, what is the primary function of the training dataset?

View answer and explanation
Question 11

Which type of machine learning involves input and output data that are labelled?

View answer and explanation
Question 12

A machine learning model that treats noise as true parameters and identifies spurious patterns is said to exhibit:

View answer and explanation
Question 13

The analysis of unstructured data in text or voice forms, such as evaluating regulatory filings, is best described as:

View answer and explanation
Question 14

Which fintech application uses computers to interpret human language, specifically for tasks like speech recognition and language translation?

View answer and explanation
Question 15

Which trading strategy specifically identifies and takes advantage of intraday securities mispricings using computers?

View answer and explanation
Question 16

Robo-advisory services typically offer portfolios with which of the following characteristics?

View answer and explanation
Question 17

What is the primary advantage of robo-advisors for customers?

View answer and explanation
Question 18

A potential disadvantage of robo-advisors during crisis periods is:

View answer and explanation
Question 19

In a distributed ledger, what mechanism is required to validate new entries?

View answer and explanation
Question 20

Which element links blocks sequentially in a blockchain?

View answer and explanation
Question 21

Computers on a blockchain network that solve cryptographic problems to validate transactions are called:

View answer and explanation
Question 22

Which type of distributed ledger network allows all participants to view all transactions and has no central authority?

View answer and explanation
Question 23

In a permissioned network, which of the following is most likely true?

View answer and explanation
Question 24

Which of the following describes an Initial Coin Offering (ICO)?

View answer and explanation
Question 25

What is a potential benefit of using distributed ledger technology for post-trade clearing and settlement?

View answer and explanation
Question 26

Electronic contracts programmed to self-execute based on agreed terms are known as:

View answer and explanation
Question 27

Tokenization refers to:

View answer and explanation
Question 28

Underfitting in a machine learning model means the model:

View answer and explanation
Question 29

Which of the following is a challenge associated with machine learning results being a 'black box'?

View answer and explanation
Question 30

Deep learning is a technique that typically uses:

View answer and explanation
Question 31

The 'Internet of Things' refers to:

View answer and explanation
Question 32

A firm has accumulated 2,000 terabytes of data. This amount is equivalent to:

View answer and explanation
Question 33

Data that are communicated periodically or with a lag are said to have:

View answer and explanation
Question 34

In the context of machine learning, 'unsupervised learning' means:

View answer and explanation
Question 35

Using ML to evaluate large volumes of research reports to detect subtle changes in sentiment is an example of:

View answer and explanation
Question 36

Which fintech application is most likely to be used for executing large orders by dividing them across exchanges?

View answer and explanation
Question 37

Cryptocurrencies typically reside on which type of network?

View answer and explanation
Question 38

A potential benefit of giving regulators permission to view a distributed ledger is:

View answer and explanation
Question 39

Mining on a blockchain requires vast resources of computing power and electricity primarily to:

View answer and explanation
Question 40

Robo-advisory services are most likely to appeal to which type of investor?

View answer and explanation
Question 41

Which data processing method involves assuring data quality by adjusting for bad or missing data?

View answer and explanation
Question 42

Which of the following is considered a 'traditional' source of data?

View answer and explanation
Question 43

An algorithm given inputs of source data with no assumptions about their probability distributions is characteristic of:

View answer and explanation
Question 44

Which of the following describes the relationship between 'volume' and 'variety' in Big Data?

View answer and explanation
Question 45

Which dataset is used to refine relationship models in machine learning?

View answer and explanation
Question 46

What is a significant drawback of Distributed Ledger Technology regarding trade cancellations?

View answer and explanation
Question 47

Investors in Initial Coin Offerings (ICOs) should be aware that:

View answer and explanation
Question 48

Which of the following is a risk analysis technique that can be enhanced by Big Data and Machine Learning?

View answer and explanation
Question 49

A key challenge in using Big Data is:

View answer and explanation
Question 50

Which technology could potentially replace paper real estate deeds at government offices?

View answer and explanation