SlideShare a Scribd company logo
1 of 8
TAO - Inria Saclay-IDF

● Machine Learning & Optimization
● Tao-uctsig:

    ● Sequential decision making

    ● One permanent, full time + others part time

    ● Applications to energy management

    ● Strong collaboration with

       ● Taiwan

       ● Artelys (Ilab Metis; joint software)

       ● Others

                               O. Teytaud, Research Fellow,
                                  olivier.teytaud@inria.fr
                                 http://www.lri.fr/~teytaud/
Power systems, high scale

                                         Production


                                         Network
Feedback mecanism
   (smart grids)




                                           em and
                                          D
For choosing investments we want
          to simulate systems
●   Difficulties:
       ●   Demand varying in time, bounded prediction
       ●   Transportation introduces constraints
       ●   Renewable ==> variability ++
●   Problems:
       ●   Limited previsibility has an impact ==> anticipative high-level
           techniques underestimate the need for storage / smoothing
       ●   Markovian assumptions ==> wrong
       ●   A system which neglects “base ≠ peak” can not be used.


    ==> Model error >> optimization error
    ==> Machine Learning on top of Math. Programming
Math programming and machine
                learning

●   Math programming:
       ●   Nearly exact solutions
       ●   High-dimensional constrained action space
       ●   But small state space & not anytime

●   Reinforcement learning
       ●   Unstable
       ●   Small / simple action space
       ●   But high dimensional state space & anytime
Stochastic dyn. Programming
                                          Huge computation time

                                           Assumes Markovian
                                                Models.

                                         Neglects non-linearities.

●   Step 1: compute Bellman's function:



                              Can work with huge constrained
●   Step 2: make decisions:            action space
Direct Policy Search
●   Define a parametric function



       ●   Neural network
       ●   Handcrafted function

●   Non-linear optimization
       ●   The best θ is the one which performs best on simulations
                 ==> obtained by non-linear stochastic optimization
       ●   Non-linearities ok, arbitrary stochastic process, large state
           space ==> little model bias
       ●   No solution for huge constrained action spaces
Math prog & reinforcement learning
●   Here, we consider “math prog = heuristic”,
    because it's fast but with strong model bias
●   Proposals:
                                                             DPS-style:
       ●   MCTS (Monte-Carlo Tree Search) + heuristic    Little model bias,
                                                         arbitrary random
       ●   DPS (Direct Policy Search) + heuristic         process, large
                                                            state space
       ●   Example              Bellman-style;
                                    ok for
                              large constrained
                                action spaces     Non-linear ~ (θ, xt)
                                                   Linear ~ x(t+1)
Works in Tao
●   Noisy non-linear optimization
      ●   Fabian's algorithm
      ●   Anytime properties (for bilevel problems)
      ●   Evolutionary algorithms
●   Reinforcement learning
      ●   MCTS (Monte Carlo Tree Search) on top of heuristics
      ●   DPS (combined with MCTS or heuristics)
●   Links with Artelys:
      ●   Joint software
      ●   Experiments
           – Non-anticipativity
           – Non-linearities

More Related Content

Viewers also liked

Combining UCT and Constraint Satisfaction Problems for Minesweeper
Combining UCT and Constraint Satisfaction Problems for MinesweeperCombining UCT and Constraint Satisfaction Problems for Minesweeper
Combining UCT and Constraint Satisfaction Problems for MinesweeperOlivier Teytaud
 
Complexity of planning and games with partial information
Complexity of planning and games with partial informationComplexity of planning and games with partial information
Complexity of planning and games with partial informationOlivier Teytaud
 
Multimodal or Expensive Optimization
Multimodal or Expensive OptimizationMultimodal or Expensive Optimization
Multimodal or Expensive OptimizationOlivier Teytaud
 
Stochastic modelling and quasi-random numbers
Stochastic modelling and quasi-random numbersStochastic modelling and quasi-random numbers
Stochastic modelling and quasi-random numbersOlivier Teytaud
 
Inteligencia Artificial y Go
Inteligencia Artificial y GoInteligencia Artificial y Go
Inteligencia Artificial y GoOlivier Teytaud
 
Noisy optimization --- (theory oriented) Survey
Noisy optimization --- (theory oriented) SurveyNoisy optimization --- (theory oriented) Survey
Noisy optimization --- (theory oriented) SurveyOlivier Teytaud
 
Tools for artificial intelligence
Tools for artificial intelligenceTools for artificial intelligence
Tools for artificial intelligenceOlivier Teytaud
 
Theories of continuous optimization
Theories of continuous optimizationTheories of continuous optimization
Theories of continuous optimizationOlivier Teytaud
 
Uncertainties in large scale power systems
Uncertainties in large scale power systemsUncertainties in large scale power systems
Uncertainties in large scale power systemsOlivier Teytaud
 
Machine learning 2016: deep networks and Monte Carlo Tree Search
Machine learning 2016: deep networks and Monte Carlo Tree SearchMachine learning 2016: deep networks and Monte Carlo Tree Search
Machine learning 2016: deep networks and Monte Carlo Tree SearchOlivier Teytaud
 
Meta Monte-Carlo Tree Search
Meta Monte-Carlo Tree SearchMeta Monte-Carlo Tree Search
Meta Monte-Carlo Tree SearchOlivier Teytaud
 
Energy Management Forum, Tainan 2012
Energy Management Forum, Tainan 2012Energy Management Forum, Tainan 2012
Energy Management Forum, Tainan 2012Olivier Teytaud
 
Dynamic Optimization without Markov Assumptions: application to power systems
Dynamic Optimization without Markov Assumptions: application to power systemsDynamic Optimization without Markov Assumptions: application to power systems
Dynamic Optimization without Markov Assumptions: application to power systemsOlivier Teytaud
 
Introduction to the TAO Uct Sig, a team working on computational intelligence...
Introduction to the TAO Uct Sig, a team working on computational intelligence...Introduction to the TAO Uct Sig, a team working on computational intelligence...
Introduction to the TAO Uct Sig, a team working on computational intelligence...Olivier Teytaud
 
Computers and Killall-Go
Computers and Killall-GoComputers and Killall-Go
Computers and Killall-GoOlivier Teytaud
 

Viewers also liked (16)

Combining UCT and Constraint Satisfaction Problems for Minesweeper
Combining UCT and Constraint Satisfaction Problems for MinesweeperCombining UCT and Constraint Satisfaction Problems for Minesweeper
Combining UCT and Constraint Satisfaction Problems for Minesweeper
 
Complexity of planning and games with partial information
Complexity of planning and games with partial informationComplexity of planning and games with partial information
Complexity of planning and games with partial information
 
Multimodal or Expensive Optimization
Multimodal or Expensive OptimizationMultimodal or Expensive Optimization
Multimodal or Expensive Optimization
 
Stochastic modelling and quasi-random numbers
Stochastic modelling and quasi-random numbersStochastic modelling and quasi-random numbers
Stochastic modelling and quasi-random numbers
 
Statistics 101
Statistics 101Statistics 101
Statistics 101
 
Inteligencia Artificial y Go
Inteligencia Artificial y GoInteligencia Artificial y Go
Inteligencia Artificial y Go
 
Noisy optimization --- (theory oriented) Survey
Noisy optimization --- (theory oriented) SurveyNoisy optimization --- (theory oriented) Survey
Noisy optimization --- (theory oriented) Survey
 
Tools for artificial intelligence
Tools for artificial intelligenceTools for artificial intelligence
Tools for artificial intelligence
 
Theories of continuous optimization
Theories of continuous optimizationTheories of continuous optimization
Theories of continuous optimization
 
Uncertainties in large scale power systems
Uncertainties in large scale power systemsUncertainties in large scale power systems
Uncertainties in large scale power systems
 
Machine learning 2016: deep networks and Monte Carlo Tree Search
Machine learning 2016: deep networks and Monte Carlo Tree SearchMachine learning 2016: deep networks and Monte Carlo Tree Search
Machine learning 2016: deep networks and Monte Carlo Tree Search
 
Meta Monte-Carlo Tree Search
Meta Monte-Carlo Tree SearchMeta Monte-Carlo Tree Search
Meta Monte-Carlo Tree Search
 
Energy Management Forum, Tainan 2012
Energy Management Forum, Tainan 2012Energy Management Forum, Tainan 2012
Energy Management Forum, Tainan 2012
 
Dynamic Optimization without Markov Assumptions: application to power systems
Dynamic Optimization without Markov Assumptions: application to power systemsDynamic Optimization without Markov Assumptions: application to power systems
Dynamic Optimization without Markov Assumptions: application to power systems
 
Introduction to the TAO Uct Sig, a team working on computational intelligence...
Introduction to the TAO Uct Sig, a team working on computational intelligence...Introduction to the TAO Uct Sig, a team working on computational intelligence...
Introduction to the TAO Uct Sig, a team working on computational intelligence...
 
Computers and Killall-Go
Computers and Killall-GoComputers and Killall-Go
Computers and Killall-Go
 

Similar to 3slides

Ilab Metis: we optimize power systems and we are not afraid of direct policy ...
Ilab Metis: we optimize power systems and we are not afraid of direct policy ...Ilab Metis: we optimize power systems and we are not afraid of direct policy ...
Ilab Metis: we optimize power systems and we are not afraid of direct policy ...Olivier Teytaud
 
Planning for power systems
Planning for power systemsPlanning for power systems
Planning for power systemsOlivier Teytaud
 
Deep Learning Tutorial
Deep Learning Tutorial Deep Learning Tutorial
Deep Learning Tutorial Ligeng Zhu
 
reinforcement learning for difficult settings
reinforcement learning for difficult settingsreinforcement learning for difficult settings
reinforcement learning for difficult settingsOlivier Teytaud
 
Distributed computing abstractions_data_science_6_june_2016_ver_0.4
Distributed computing abstractions_data_science_6_june_2016_ver_0.4Distributed computing abstractions_data_science_6_june_2016_ver_0.4
Distributed computing abstractions_data_science_6_june_2016_ver_0.4Vijay Srinivas Agneeswaran, Ph.D
 
Challenges in Large Scale Machine Learning
Challenges in Large Scale  Machine LearningChallenges in Large Scale  Machine Learning
Challenges in Large Scale Machine LearningSudarsun Santhiappan
 
Towards quantum machine learning calogero zarbo - meet up
Towards quantum machine learning  calogero zarbo - meet upTowards quantum machine learning  calogero zarbo - meet up
Towards quantum machine learning calogero zarbo - meet upDeep Learning Italia
 
Introduction to deep learning workshop
Introduction to deep learning workshopIntroduction to deep learning workshop
Introduction to deep learning workshopShamane Siriwardhana
 
Reinforcement learning
Reinforcement learningReinforcement learning
Reinforcement learningDongHyun Kwak
 
Smartphone Activity Prediction
Smartphone Activity PredictionSmartphone Activity Prediction
Smartphone Activity PredictionTriskelion_Kaggle
 
Optimization of power systems - old and new tools
Optimization of power systems - old and new toolsOptimization of power systems - old and new tools
Optimization of power systems - old and new toolsOlivier Teytaud
 
Tools for Discrete Time Control; Application to Power Systems
Tools for Discrete Time Control; Application to Power SystemsTools for Discrete Time Control; Application to Power Systems
Tools for Discrete Time Control; Application to Power SystemsOlivier Teytaud
 
Optimizing Deep Networks (D1L6 Insight@DCU Machine Learning Workshop 2017)
Optimizing Deep Networks (D1L6 Insight@DCU Machine Learning Workshop 2017)Optimizing Deep Networks (D1L6 Insight@DCU Machine Learning Workshop 2017)
Optimizing Deep Networks (D1L6 Insight@DCU Machine Learning Workshop 2017)Universitat Politècnica de Catalunya
 
Jay Yagnik at AI Frontiers : A History Lesson on AI
Jay Yagnik at AI Frontiers : A History Lesson on AIJay Yagnik at AI Frontiers : A History Lesson on AI
Jay Yagnik at AI Frontiers : A History Lesson on AIAI Frontiers
 
Big data 2.0, deep learning and financial Usecases
Big data 2.0, deep learning and financial UsecasesBig data 2.0, deep learning and financial Usecases
Big data 2.0, deep learning and financial UsecasesArvind Rapaka
 
Optimization Heuristics
Optimization HeuristicsOptimization Heuristics
Optimization HeuristicsKausal Malladi
 

Similar to 3slides (20)

Ilab Metis: we optimize power systems and we are not afraid of direct policy ...
Ilab Metis: we optimize power systems and we are not afraid of direct policy ...Ilab Metis: we optimize power systems and we are not afraid of direct policy ...
Ilab Metis: we optimize power systems and we are not afraid of direct policy ...
 
Planning for power systems
Planning for power systemsPlanning for power systems
Planning for power systems
 
Deep Learning Tutorial
Deep Learning Tutorial Deep Learning Tutorial
Deep Learning Tutorial
 
Power systemsilablri
Power systemsilablriPower systemsilablri
Power systemsilablri
 
reinforcement learning for difficult settings
reinforcement learning for difficult settingsreinforcement learning for difficult settings
reinforcement learning for difficult settings
 
Space time & power.
Space time & power.Space time & power.
Space time & power.
 
Distributed computing abstractions_data_science_6_june_2016_ver_0.4
Distributed computing abstractions_data_science_6_june_2016_ver_0.4Distributed computing abstractions_data_science_6_june_2016_ver_0.4
Distributed computing abstractions_data_science_6_june_2016_ver_0.4
 
Challenges in Large Scale Machine Learning
Challenges in Large Scale  Machine LearningChallenges in Large Scale  Machine Learning
Challenges in Large Scale Machine Learning
 
Towards quantum machine learning calogero zarbo - meet up
Towards quantum machine learning  calogero zarbo - meet upTowards quantum machine learning  calogero zarbo - meet up
Towards quantum machine learning calogero zarbo - meet up
 
Introduction to deep learning workshop
Introduction to deep learning workshopIntroduction to deep learning workshop
Introduction to deep learning workshop
 
Reinforcement learning
Reinforcement learningReinforcement learning
Reinforcement learning
 
Smartphone Activity Prediction
Smartphone Activity PredictionSmartphone Activity Prediction
Smartphone Activity Prediction
 
Optimization of power systems - old and new tools
Optimization of power systems - old and new toolsOptimization of power systems - old and new tools
Optimization of power systems - old and new tools
 
Tools for Discrete Time Control; Application to Power Systems
Tools for Discrete Time Control; Application to Power SystemsTools for Discrete Time Control; Application to Power Systems
Tools for Discrete Time Control; Application to Power Systems
 
Optimizing Deep Networks (D1L6 Insight@DCU Machine Learning Workshop 2017)
Optimizing Deep Networks (D1L6 Insight@DCU Machine Learning Workshop 2017)Optimizing Deep Networks (D1L6 Insight@DCU Machine Learning Workshop 2017)
Optimizing Deep Networks (D1L6 Insight@DCU Machine Learning Workshop 2017)
 
Jay Yagnik at AI Frontiers : A History Lesson on AI
Jay Yagnik at AI Frontiers : A History Lesson on AIJay Yagnik at AI Frontiers : A History Lesson on AI
Jay Yagnik at AI Frontiers : A History Lesson on AI
 
Deep learning - a primer
Deep learning - a primerDeep learning - a primer
Deep learning - a primer
 
Deep learning - a primer
Deep learning - a primerDeep learning - a primer
Deep learning - a primer
 
Big data 2.0, deep learning and financial Usecases
Big data 2.0, deep learning and financial UsecasesBig data 2.0, deep learning and financial Usecases
Big data 2.0, deep learning and financial Usecases
 
Optimization Heuristics
Optimization HeuristicsOptimization Heuristics
Optimization Heuristics
 

3slides

  • 1. TAO - Inria Saclay-IDF ● Machine Learning & Optimization ● Tao-uctsig: ● Sequential decision making ● One permanent, full time + others part time ● Applications to energy management ● Strong collaboration with ● Taiwan ● Artelys (Ilab Metis; joint software) ● Others O. Teytaud, Research Fellow, olivier.teytaud@inria.fr http://www.lri.fr/~teytaud/
  • 2. Power systems, high scale Production Network Feedback mecanism (smart grids) em and D
  • 3. For choosing investments we want to simulate systems ● Difficulties: ● Demand varying in time, bounded prediction ● Transportation introduces constraints ● Renewable ==> variability ++ ● Problems: ● Limited previsibility has an impact ==> anticipative high-level techniques underestimate the need for storage / smoothing ● Markovian assumptions ==> wrong ● A system which neglects “base ≠ peak” can not be used. ==> Model error >> optimization error ==> Machine Learning on top of Math. Programming
  • 4. Math programming and machine learning ● Math programming: ● Nearly exact solutions ● High-dimensional constrained action space ● But small state space & not anytime ● Reinforcement learning ● Unstable ● Small / simple action space ● But high dimensional state space & anytime
  • 5. Stochastic dyn. Programming Huge computation time Assumes Markovian Models. Neglects non-linearities. ● Step 1: compute Bellman's function: Can work with huge constrained ● Step 2: make decisions: action space
  • 6. Direct Policy Search ● Define a parametric function ● Neural network ● Handcrafted function ● Non-linear optimization ● The best θ is the one which performs best on simulations ==> obtained by non-linear stochastic optimization ● Non-linearities ok, arbitrary stochastic process, large state space ==> little model bias ● No solution for huge constrained action spaces
  • 7. Math prog & reinforcement learning ● Here, we consider “math prog = heuristic”, because it's fast but with strong model bias ● Proposals: DPS-style: ● MCTS (Monte-Carlo Tree Search) + heuristic Little model bias, arbitrary random ● DPS (Direct Policy Search) + heuristic process, large state space ● Example Bellman-style; ok for large constrained action spaces Non-linear ~ (θ, xt) Linear ~ x(t+1)
  • 8. Works in Tao ● Noisy non-linear optimization ● Fabian's algorithm ● Anytime properties (for bilevel problems) ● Evolutionary algorithms ● Reinforcement learning ● MCTS (Monte Carlo Tree Search) on top of heuristics ● DPS (combined with MCTS or heuristics) ● Links with Artelys: ● Joint software ● Experiments – Non-anticipativity – Non-linearities