Most of the energy and attention in machine learning focused on the model training side of the problem. Multiple frameworks, in every language, provide developers with access to a host of data manipulation and training algorithms, but until recently developers had virtually no frameworks to build out predictive engines from trained ML models. Most developers resorted to building custom applications, but building highly available, highly performant applications is difficult. Redis in conjunction with the Redis-ML module provides a server framework for developers to build predictive engines with familiar, off-the-shelf components. Developers can take advantage of all the features of Redis to deliver faster and more reliable prediction engines with less custom development. This talk is a technical session which examines how Redis can be used in conjunction with a Spark based training platform to deliver real-time predictive and decision making features as part of a larger system. To set the context for the session, we start with an introduction to the Redis data model and how features of Redis (namespace, replication) can be used to build fast predictive engines (at scale), that are more reliable, more feature rich and easier to manage than custom applications. From there, we look at the model serving capabilities of Redis-ML and how they can be integrated with a Spark-based ML pipeline to automate the entire model development process from training to deployment. The session ends with a demonstration of a simple machine learning pipeline. Using Spark we train several example models, load them directly into Redis and demonstrate Redis as the predictive engine for making real-time recommendations. At the end of the session, developers should feel confident that they could use Redis as a server framework to build a predictive serving engine for a Spark-based ML pipeline.

1. Tague Griffith, Redis Labs
Deploying Real-Time Decision
Sevices using Redis
#MLSAIS12

2. Why Machine Learning

3. Teaching a computer, by example, an
algorithm that is too complex to program

4. Machine Learning Problems
Pick One of a Set
• Spam Detection
• Manufacturing defect
detection
• Handwriting analysis
• Decision Trees
• Naïve Bayes
• Logistic Regression
Score or Rank
• Recommendations
• Likelihood of
Purchase
• Linear Regression
• SVM
Classification Regression
Group Similar
• Find Similar Items
• Customer
segmentation
• Cohort detection
• K-Means
• K-Nearest Neighbors
• Hierarchical
Clustering
Clustering

5. Supervised Learning – Training Spam
Classifier
#MLSAIS12

6. Deploying a Spam Classifier
#MLSAIS12

7. How do we Build these Boxes
¯_( )_/¯
#MLSAIS12

8. • Building high performance and reliable
services are hard, isn't there something
we can deploy

9. Redis - ML

10. Typical Spark Application Structure
Spark Training
Data is loaded into Spark Model is saved in files
File System Custom Server
Model is loaded to your
custom app
Serving Client
Client App
#MLSAIS12

11. Redis-ML: Predictive Model Serving Engine
• Predictive models as native Redis types
• Perform evaluation directly in Redis
• Store training output as “hot model”
Spark Training
Data loaded into Spark Model is saved in
Redis-ML
Redis-ML
Serving Client
Client
App
Client
App
Client
App
Any Training
Platform

12. REmote DIctionary Server
Strings Hashes Lists
Sets Bitmaps
Hyperlog-
logs
Sorted
Sets
Geo-
spatial
Bitfield

13. A Quick Recap of Redis
Key
"I'm a Plain Text String!"
{ A: “foo”, B: “bar”, C: “baz” }
Strings / Bitmaps / BitFields
Hash Tables (objects!)
Linked Lists
Sets
Sorted Sets
Geo Sets
HyperLogLog
{ A , B , C , D , E }
[ A → B → C → D → E ]
{ A: 0.1, B: 0.3, C: 100, D: 1337 }
{ A: (51.5, 0.12), B: (32.1, 34.7) }
00110101 11001110 10101010

14. Redis Modules
• Any C/C++ program can now run on Redis
• Use existing or add new data-structures
• Enjoy simplicity, infinite scalability and high availability while
keeping the native speed of Redis
• Can be created by anyone
New Capabilities
New Commands
New Data Types

15. Redis ML Module
Redis Module
Tree Ensembles
Linear Regression
Logistic Regression
Matrix + Vector Operations
More to come...

16. Random Forest Model
• A collection of decision trees
• Supports classification & regression
• Splitter Node can be:
◦ Categorical (e.g. day == “Sunday”)
◦ Numerical (e.g. age < 43)
• Decision is taken by the majority of decision trees

17. Classic Tree Problem: Titanic Survival
YES
Sex =
Male ?
Age <
9.5?
Sibps >
2.5?
Survived
Died
SurvivedDied
NO
• Passenger Data encoded as feature vecto
• ML Algorithm learns the tree rules
• ID3, CART (RPART), etc.
• Tree rules used to infer results

18. Titanic Survival: Random Forest
YES
Sex =
Male ?
Age <
9.5?
*Sibps >
2.5?
Survived
Died
SurvivedDied
NO YES
Country=
US?
State =
CA?
Height>
1.60m?
Survived
Died
SurvivedDied
NO YES
Weight<
80kg?
I.Q<100?
Eye color
=blue?
Survived
Died
SurvivedDied
NO
Tree #1 Tree #2 Tree #3

19. Who Would Survive the Titanic
• John:
• Male, 34,
• Married w/ 2 kids
(Sibps=3)
• New York, USA
• 1.78m, 78kg
• 110 iq
• Blue eyes
Mathew:
• Male, 6
• 3 Sisters (Sibps=3)
• New York, USA
• 1.06m, 22.7 kg
• 100 iq
• Brown eyes
Let's use our forest to find out

20. Redis: Forest Data Type
Add nodes to a tree in a forest:
Perform classification/regression of a feature vector:
ML.FOREST.ADD <forestId> <treeId> <path>
[ [NUMERIC|CATEGORIC] <splitterAttr> <splitterVal> ] |
[LEAF] <predVal>
ML.FOREST.RUN <forestId> <features>
[CLASSIFICATION|REGRESSION]

21. Real World Challenge
• Ad serving company
• Need to serve 20,000 ads/sec @ 50msec data-center latency
• Runs 1k campaigns → 1K random forest
• Each forest has 15K trees
• On average each tree has 7 levels (depth)

22. Ad Serving costs: Homegrown v. Redis
Homegrown
1,247 x c4.8xlarge 35 x c4.8xlarge
Cut computing infrastructure
by 97%

23. Summary
• Train with Spark, Serve with Redis
• 97% resource cost serving
• Simplify ML lifecycle
• Redise (Cloud or Pack):
‒Scaling, HA, Performance
‒PAYG – cost optimized
‒Ease of use
‒Supported by the teams who created Spark and
Redis
Spark Training
Data loaded into Spark Model is saved in
Redis-ML
Redis-ML
Serving Client
Client
App
Client
App
Client
App
+

24. Thank you!