[VFS 2019] AI in Finance

[VFS 2019] AI in Finance

• 1.
  Artificial Intelligence in Finance LeCong Binh, MBA Vietnam Frontier Summit 2019
• 2.
  LE CONG BINH,MBA Applied Mathematics and Informatics, VNU-HUS MBA, NEU Advanced Data Science with IBM 15+ years in banking industry 10+ Data warehouse, 3+ Data mining experience [email protected]
• 3.
  Data science methodology Datascience algorithm Use case in banking sector CRISP-DM: cross-industry process for data mining Time series, Machine and Deep learning Risk, sales and customer insight TABLE OF CONTENTS 01 02 03 Want to hear your question 04 Discussion
• 4.
  Harvard Business Review “DataScientist the sexiest job of the 21st Century ”
• 5.
  Foundational methodology Business acumen Communicationskills Data processing Data visualization Mathematical Statistical Machine Learning Deep Learning Programing
• 6.
  DATA VISUALIZATION Business acumen Communicationskills Data processing Data visualization Mathematical Statistical Machine Learning Deep Learning Programing “The most of time in data science is Collect, Cleaning and Organizing Data”
• 7.
  Business acumen Communication skills Dataprocessing Data visualization Mathematical Statistical Machine Learning Deep Learning Programing “Python is the top of programing language”
• 8.
  DATA VISUALIZATION Business acumen Communicationskills Data processing Data visualization Mathematical Statistical Machine Learning Deep Learning Programing Data Visualizations make big and small data easier for the human brain to understand, and visualization also makes it easier to detect patterns, trends, and outliers in groups of data
• 9.
  Data Science Algorithm 02 Input +Rules => Target Input + Target => Rules
• 10.
  ALGORITHYM AR MASMA ARIMA ARIMAX VAR VARMA A series of values of a quantity obtained at successive times, often with equal intervals between them Holt Winter’s Exponential
• 11.
  ALGORITHYM Supervised Unsupervised Reinforcement Data isthe “new oil” that intelligent algorithms consume: the more data is given in input, the more accurate the prediction output is.
• 12.
  ALGORITHYM Algorithms tree
• 13.
  DEEP LEARNING FRAMEWORK Convolutionnetwork Sequential RNN LSTM /GRU RBMs Neutral Epoch Dense Layer
• 14.
  LONG SHORT TERMMEMORY
• 15.
  15 MODEL EVALUATION ESTIMATORS = { 'KernelRidge':KernelRidge(alpha=1.0), 'BayesianRidge': BayesianRidge(compute score=True), 'SVR': svm.SVR(), 'GPR': GaussianProcessRegressor() } X train, X test, y train, y test, X pred = CreateTrainTestDataset()
• 16.
  Banking use case03 Bydesign, intelligent algorithms are good at solving specific problems and cannot deviate from what they were designed for. An algorithm trained to detect suspicious payments would not be able to detect any other suspicious activity related to trading, for instance. In addition, algorithms are purely rational and lack essential factors such as emotional intelligence and the ability to contextualize information, unlike human beings. That’s why banking Chabot's often disappoint: they are “smart” but lack empathy.
• 17.
  USE CASE Source: https://emerj.com/ai- sector-overviews/ai- in-banking-analysis/
• 18.
  BASIC USE CASE
• 19.
  USE CASE
• 20.
  SUMMARY 1. Financial institutionsmay need to consider restructuring their business models in order to harness this new technology. 2. There will be a need to create job positions in relation to AI 3. However, proceed with some caution due to the existing challenges in terms of trust, privacy and data issues. 4. They must also keep an on RegTech (Regulatory Technology) to better address new regulatory requirement. 5. Overall, firms should embrace AI now or face the possibility of being left behind
• 21.
  21 FOR BEGINER Roadmap reference: https://medium.com/ml-research-lab/data-scientist- learning-path-2018-a82e67d49d8e Online Training: https://www.coursera.org/professional- certificates/ibm-data-science https://cognitiveclass.ai/learn/data-science
• 22.
  Does anyone haveany questions? [email protected] THANKS!