AI in finance - from hype to marketing and cybersecurity use cases

1
Natalino Busa - @natbusa
Global Artificial Intelligence
Conference
AI in Finance:
from Hype to marketing and
cyber security use cases

www.globalbigdataconference.com
Twitter : @bigdataconf
Global Artificial Intelligence Conference
AI in Finance:
from Hype to marketing and cyber security use cases
Natalino Busa
Twitter : @natbusa

3 Natalino Busa - @natbusa
Cognitive Finance Group Advisory Board Member
ING Group Enterprise Architect: Cybersecurity, Fintech
Teradata Head of Applied Data Science
Teradata Global Evangelist on Open Sourced Technologies
O’Reilly Author and Speaker
Philips Senior Researcher, Data Architect
Linkedin and Twitter:
@natbusa

4
What about AI in Finance?

5
The Medici Bank:
Italian: Banco Medici
1397–1494

6
Data as a Relationship
● Trust
● Transparency of Use
● Customer First
● Regulations and Laws
● Respect and Protect
● Providing a Service

An ethical approach
for Actionable Financial Data
Help the customer
Propose, Advise, Select, Filter, Connect,
Simplify1.
Protect the customer
Detect, Prevent, Alert, Block, Defend,
Identify, Authorize
2.

8
Personalized Financial

http://www.slideshare.net/ING/4q15-media
● Innovation helps to empower people to make better
financial decisions. ING, has launched several new
omni-channel banking platforms.
● The platform gives customers insights
into their personal finances in an easy
and intuitive way.
Financial personalized recommenders

Financial personalized recommenders
● It Knows Finance
● Conversational
● Personal
● Actionable
● Predictive
● Reuse Existing Content

Inspiration from the Web

12
Credit Pre-Authorization

● Fintech innovation to help strengthen our lending
capabilities and better serve our consumer and SME
clients.
● Kabbage, one of the leading US-based technology
platforms providing automated lending to SME.
● In January 2016, ING has made an investment in
fintech WeLab, which provides consumer loans in
China and Hong Kong in a fully automated process
that just takes minutes, from application to approval.
http://www.slideshare.net/ING/4q15-media
Strategic data-driven initiatives

14
Approaching (Almost) Any Machine Learning Problem
- Abhishek Thakur, Kaggle Grandmaster -
data labels
raw data: tables, files Useful dataData munging Feature
Engineering
Tabular Data ready for ML

Input
Hand Designed
Program
Input Input
Rule-based System
Output
Hand Designed
Features
Mapping from
features
Output
Learned
Features
Mapping from
features
Output
Classic Machine
Learning
Input
Learned
Features
Learned
Complex features
Output
Mapping from
features
Representational
Machine Learning
Deep Learning
(end-to-end learning)
Prof. Yoshua Bengio - Deep Learning
https://youtu.be/15h6MeikZNg
Predictive API’s: How to get there?

From Feature to Architecture Engineering:

Demo:
Credit Payment Defaulting
with TensorFlow and Keras
Methodology
This research aimed at the case of customers
default payments in Taiwan and compares the
predictive accuracy of probability of default among
six data mining methods. From the perspective of
risk management, the result of predictive
accuracy of the estimated probability of default
will be more valuable than the binary result of
classification - credible or not credible clients https://archive.ics.uci.edu/ml/datasets/default+of+credit+card+clients

Step 0: data exploration
Target variable: default payment next month
Color scheme: yes, defaulting not defaulting g

Step 1: feature engineering
pay_1 -1
pay_2 0
pay_3 -1
pay_4 0
pay_5 0
pay_6 0
pay_avgamt1 0.203221
pay_avgamt2 3.72718
pay_avgamt3 1.01611
pay_stdavgamt 1.40083
pay_avg -0.333333
pay_std 0.516398

Step 1: baseline (e.g regression)
model = Sequential()
model.add(Dense(1, input_shape=(input_dim,))
model.add(Activation('relu'))
1
87
it’s a neural network …
with no network :)

Step 2: deep learning
model = Sequential()
model.add(Dense(256, input_shape=(input_dim,), activation='relu'))
model.add(Dense(256, activation='relu'))
model.add(Dropout(0.25))
model.add(Dense(10, activation='sigmoid'))
model.add(Dense(1))
model.add(Activation('sigmoid'))
256
64
64
87
256
64
64
10
1

Step 3: compare: is deep learning better?
256
64
64
87
256
64
64
10
1
1
87
Shallow
Logit Model
Deep
Learning

Step 4: picking the brain of our DL model
87
1

256
64
64
87
256
64
64
10
1
Step 4: picking the brain of our DL model

Step 5: semantic clustering
Default
Very Safe
Mixed
Group
Safe
SafeMixed
Group

26
Hands on with Keras and Tensorflow

27
Hyper-Parameters tuning
- based on scikit-learn
- 15 classifiers,
- 14 feature preprocessing methods
- 4 data preprocessing methods
- 110 hyperparameters
- Supervised classification challenge:
100 different datasets
https://arxiv.org/abs/1611.03824v1

The API for banking data.
Two levels:
- Transactions
- Risk Scoring
Inspiration from the Web

29
Card Theft: Geo-Alerting

Clustering geolocated data
using Spark and DBSCAN
How to group users’ events using machine learning and distributed computing
By Natalino Busa
Predictive API’s: Clustering Geolocated Data

@natbusa | linkedin.com: Natalino Busa
Venues and Events

@natbusa | linkedin.com: Natalino BusaEvents clustering

Card Theft/Cloning: DBSCAN and Convex Hulls

Fast writes
2D Data Structure
Replicated
Tunable consistency
Multi-Data centers
CassandraKafka Spark
Streaming Events
Distributed, Scalable Transport
Events are persisted
Decoupled Consumer-Producers
Topics and Partitions
Ad-Hoc Queries
Joins, Aggregate
User Defined Functions
Machine Learning,
Advanced Stats and Analytics
Kafka+Cassandra+Spark: SMACK stack
Streaming Machine Learning

Spark: Unified Distributed Computing:
SQL + Machine Learning + Graph Analytics
Spark - RDDs
Streaming SQL MLlib Graphx
Analytics, Statistics, Data
Science, Model Training
HDFS NoSQL SQL
Data Sources
Map-Reduce
HDFS KAFKA
Hive

Cassandra: Store all the data
Spark: Analyze all the data
DC1: replication factor 3 DC2: replication factor 3 DC3: replication factor 3 + Spark Executors
Storage! Analytics!
Data
Spark and Cassandra: distributed goodness

Cassandra - Spark Connector
Cassandra: Store all the data
Spark: Distributed Data Processing
Executors and Workers
Cassandra-Spark Connector:
Data locality,
Reduce Shuffling
RDD’s to Cassandra Partitions
DC3: replication factor 3 +
Spark Executors

38
Cyber security in Finance

39
Network Intrusion Detection
It contains 130 million flow records involving
12,027 distinct computers over 36 days (not
the full 58 days claimed for the entire data
release).
Each record consists of: time (to nearest
second), duration, source and destination
computer ids, source and destination ports,
protocol, number of packets and number of
bytes
Techniques: TDA, Dimensionality Reduction
https://en.wikipedia.org/wiki/Nonlinear_dimensionality_reduction

AI: tools and technologies

Tools for AI and Machine (deep) Learning
… this are just a few examples ...

AI: models and algorithms

AI: an ensemble of analytical methods
SQL + Graph + Text + Machine Learning + Voice/Image/Video

44
AI in Finance:
Recap & Lessons Learned

45
Takeaways
● AI can be applied in Finance: YES
● Train your AI: Domain Experts + ML
● Use All Tools, All Data

Distributed computing Artificial Intelligence
Machine Learning Statistics Big/Fast Data
Streaming Computing
Linkedin and Twitter:
@natbusa

AI in finance - from hype to marketing and cybersecurity use cases

Recommandé

Recommandé

Contenu connexe

Plus de Natalino Busa

Plus de Natalino Busa (17)

Dernier

Dernier (20)

AI in finance - from hype to marketing and cybersecurity use cases