Se ha denunciado esta presentación.
Utilizamos tu perfil de LinkedIn y tus datos de actividad para personalizar los anuncios y mostrarte publicidad más relevante. Puedes cambiar tus preferencias de publicidad en cualquier momento.

Accelerating Machine Learning with Cognitive Calibration - Kalpesh Balar, Coseer

272 visualizaciones

Publicado el

Machine learning is one of the most prominent components of technology and development. See how cognitive calibration accelerates machine learning along with the benefits, challenges, and products.

Kalpesh Balar, Coseer

Publicado en: Tecnología
  • Inicia sesión para ver los comentarios

  • Sé el primero en recomendar esto

Accelerating Machine Learning with Cognitive Calibration - Kalpesh Balar, Coseer

  1. 1. A C C E L E R A T I N G M A C H I N E L E A RN I N G W I T H kalpesh@coseer.com www.coseer.com 0
  2. 2. COSEER 1 Web/ XML Pages Social Media Third Party Research Proprietary Databases Internal Documents Automated Workflows (Tasks/ Processes/ Decisions)
  3. 3. EXAMPLES • Building healthcare product database from 35m documents • Compiling 3m documents every day into actionable stock insights • Reading through complex legal/ technical documents to answer questions 2
  4. 4. TACTICAL COGNITIVE COMPUTING 3
  5. 5. 4 About Cognitive Calibration Implementation Case Study
  6. 6. COGNITIVE CALIBRATION Qualify data based on its value + Train differentially 5
  7. 7. 6 INCREASED ACCURACY
  8. 8. LOWER COMPUTE INTENSITY 7
  9. 9. ABILITY TO MANAGE OUTLIERS 8 Sky Color 1200 PST San Francisco Sky Color ?? 1200 LTST Mars Pathfinder
  10. 10. 9 About Cognitive Calibration Implementation Case Study
  11. 11. TIERED TRAINING 10 Constrain Solution Space Constrain Solution Space TrainFilter Filter Reliability: High Medium Low
  12. 12. CALIBRATION OF INPUTS 11 Cognitive Calibration Classifier Reliability: High Medium Low
  13. 13. METADATA CLASSIFIER Map: m  Reliability Tier • Examples – Source, Author, Format (image/ text) 12
  14. 14. DETERMINISTIC CLASSIFIER Map: f(x)  Reliability Tier • Source order as per the context • Completeness • Grammatical accuracy 13
  15. 15. PROBABILISTIC CLASSIFIER f(x)  p(R) • Probabilistic distribution of the tier • f(x) can be machine learnt independently e.g. political correctness of statements 14
  16. 16. PROBABILISTIC CLASSIFIER 15 Sculpt Solution Space Sculpt Solution Space Train Probability Distribution Probability Distribution Reliability: High Medium Low
  17. 17. CLOSED LOOP CLASSIFIER F(RO)  F( f(x)  p(R) ) • Probability distribution of a Probabilistic Classifier is improved by results of the main machine • Closed Loop Classifier  Deterministic Classifier 16
  18. 18. CLOSED LOOP CLASSIFIER 17 Probabilistic Classifier Primary Trainer Reliability: High Medium Low Continuous Learning
  19. 19. 18 About Cognitive Calibration Implementation Case Study
  20. 20. PRODUCT DATABASE Client Problem 19 • One of the largest healthcare Companies • 10m+ SKUs • No standardized database comparing attributes of products • Previous human attempts to build database unsuccessful
  21. 21. INPUT • Product Brochures • White Papers • Surgical Protocols • Sporadic human entries by previous attempts 20 Drill Bit 4.5 mm Cannulated Jacobs Chuck With 135 mm Stop 165 mm Humeral Nail- EX Instrument 03.010.089 03.010.089 DRILLBIT 2.0 MM GUIDE WIRE 4.5MM CANNULATED DRILL BIT JC/WITH 135MM STOP/165MM 45mm cann BIT BIT DRILL 03.010.089* BIT DRILL CANN 4.5MM X 165MM BIT DRL CANN 4.5X135MM BIT DRL CNLD 4.5X165MM DRILL Drill Bit 4.5 mm Cnltd Jacobs Chuck 135mm…
  22. 22. CHALLENGES • Erroneous data e.g. human entries • Different levels of detail e.g. Metal, CoCr, CoCr46 • Extensive use of context specific shorthand e.g. OD = outer diameter for acetabular shells, OD = optical density for intraocular lenses • Incomplete or missing data 21
  23. 23. MANAGING CONFLICTING DATA • Sources tiered based on accuracy and specificity Product Brochures White Papers/ Surgical Protocols Authenticated Web Data Human entries in the system Non-authenticated web and other data 22
  24. 24. IDENTIFYING MATERIALS • Confusion of materials of the part itself and corresponding parts. – e.g. “Coated Tube Poly Silicone” – Unclear if coating is Poly or Tube is Poly • Constrained solution spaces using Deterministic Classifiers help – e.g. “parts use plastic tubes” tubes cannot be silicone. 23
  25. 25. FIGURING OUT SHORT HANDS 24 Probabilistic Classifier • Classifies docs if other info identifies part category • Trainer uses high reliability data to identify more classifying features Primary Trainer Reliability: High All corpus “outer diameter” “optical density” Continuous Learning “outer diameter” “optical density” Reliability: High Reliability: Low “outer diameter” “optical density”
  26. 26. OUTPUT • Learnt attributes and their corresponding values for all parts • Impossible without multiple cognitive calibration frameworks deployed in models • Difficult to implement without other strengths in tactical cognitive computing 25 Category Sub-Category Diameter Stop Length Guide Wire Dia Cannulated Radiolucent Coupling Platform Reusable Drill Bit Humeral Nail 4.5 mm 135.0 mm 165.0 mm 2.0 mm Yes No Jacobs Chuck Synthes (Estd.) Yes
  27. 27. T H A N K S kalpesh@coseer.com www.coseer.com 26

×