Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
Upcoming SlideShare
Webinar - Know Your Customer - Arya (20160526)
Next
Download to read offline and view in fullscreen.

0

Share

Download to read offline

Webinar - Patient Readmission Risk

Download to read offline

presented by Neel Kishan

Related Books

Free with a 30 day trial from Scribd

See all

Related Audiobooks

Free with a 30 day trial from Scribd

See all
  • Be the first to like this

Webinar - Patient Readmission Risk

  1. 1. 11 Dato Confidential - Do not Distribute1 Neel Kishan, Technical Sales Lead neel@dato.com Building Applications to Assess Patient Readmission Risk
  2. 2. Hello my name is Neel Kishan Technical Sales Lead (former neuroscientist, GPU programmer, Eagle Scout, Chicago sports fan) 2 neel@dato.com Let’s Schedule a Time to Talk: https://calendly.com/dato-neel
  3. 3. Poll: Getting to know you 1. What do you do? 2. Are you using a data-driven approach to reducing readmissions today? 3
  4. 4. Why we are here today – Reducing Hospital Readmissions TheProblem Patient care requires innovative methods to address the complexity for improving outcomes Readmission rates exceed 17% and most of these are avoidable Medicare spends $17B for avoidable readmissions CurrentSituation Hospitals have started to use analytics such as the LACE index to decrease readmission rates The Readmission Reduction Program (HRRP) reduces payments up to 3% for hospitals with excess readmissions for specific diagnoses NeedforReal-timeInsight Most analytic tools are not specific and do not leverage the wealth of data stored in EMRs, including text, numeric, and image data. Predictive risk scoring need to be explainable to all healthcare professionals 4
  5. 5. Methods for Understanding Readmission Risk Difficulty of Implementation Precision Intuition • Health care professionals are experts who understand emergent phenomena • Like all humans, prone to blind spots Analytic Approach • Rules based approaches provide recommendations on data • They do not provide actionable insights Machine Learning • Can learn from highly complex data and self organize to understand risk • Provides real-time feedback to healthcare professionals • Analyzes the efficacy of proactive measures 6
  6. 6. Precise, Data Driven Healthcare Requires Machine Learning • Data Quality Analysis • Precision Medicine • Radiology Image Analysis • Fraud, Waste, and Abuse • Connected Devices • Clinical Decision Support 7
  7. 7. Dato: The Platform for Real-Time Machine Learning 8
  8. 8. 8 Dato’s Machine Learning Core Tenets • Maps business tasks to machine learning routines • Eliminates bottlenecks to production • Simplifies iteration & understanding Create Value Fast • Easily combine any variety of features & ML tasks with any data • Platform components are open, reusable, & sharable • Easily extend & integrate with other frameworks Flexibility to Innovate • Make ML safe & consumable for the enterprise • Easily deploy, manage, and improve ML as intelligent micro-services • Adapt to a changing world that drifts from your historical data Intelligence in Production
  9. 9. Dato products 9 Dato Confidential - Do not Distribute
  10. 10. 10 Dato’s Deep ML Capabilities Application Toolkits • Auto-select the best algorithm • Auto-feature engineering for task • App-centric visualizations Robust Enterprise-Grade Algorithms • 50+ of best-practice & novel algorithms • Robust to real-world data 181#secs# 266#secs# 544#secs# # Dato#(10node)# Spark#(50Node)# # Vowpal#Wabbit# Time#(s)# Matrix factorization PageRank 0 2000 4000 0 4 8 12 16 Runtime(s) #Machines Criteo (4B rows) Logistic Regression Common Crawl (100B rows)Netflix (100M rows) Only platform with scalable Deep Learning, Boosted Trees, Graph Analytics, & more
  11. 11. Dato Predictive Services GraphLab Create/ Dato Distributed Rapid model building Deploy as microservice Live serving, monitoring, & model management Iterate and improve on your infrastructure: How Dato Makes Data Science Agile for Organizations Dato Confidential - Do not Distribute11
  12. 12. Dato Products - The Agile Machine Learning Platform Dato Confidential - Do not Distribute12
  13. 13. Poll: Data Science at your workplace 1. Does your team have data scientists or software developers? 2. Are you using Machine Learning in production today? 13
  14. 14. Readmission Scoring: Machine Learning Process Supervised Machine Learning workflow: Historical Data • Split train/test datasets • Readmissions& non- readmissions Train ML Model • Use the medical history of patients • Use interaction of patients Deploy • Predict likelihood to be readmitted to hospital 14
  15. 15. Using Dato to Predict Early Readmission Based on 100,000 patient interactions Demo 15
  16. 16. Explanation Advanced Readmission Risk Applications in Production 0 100 Intercranial Pressure Lab Result Saturation Automatic Feature Extraction Medical History, Labs, Procedures Automatic Feature Extraction Risk Score Advanced ML model Provider Network Relationships Intelligent Application Patient-Provider Data 16
  17. 17. Thank you! Want to find out how to incorporate machine learning into your organization? Ping me email: neel@dato.com Or Visit Us at the Data Science Summit http://bit.ly/DSS-SF-2016 Discount Code: DSSFriend

presented by Neel Kishan

Views

Total views

411

On Slideshare

0

From embeds

0

Number of embeds

0

Actions

Downloads

21

Shares

0

Comments

0

Likes

0

×