Intro; Preface; Contents; Acronyms; 1 Introduction; 1.1 Motivation; 1.2 Goals; 1.3 Book Outline; References; 2 Background; 2.1 Time Series Analysis; 2.1.1 Euclidean Distance; 2.1.2 Dynamic Time Warping; 2.1.3 Piecewise Linear Approximation; 2.1.4 Piecewise Aggregate Approximation; 2.1.5 Symbolic Aggregate approXimation; 2.2 Genetic Algorithm; 2.2.1 Selection Operator; 2.2.2 Crossover Operator; 2.2.3 Mutation Operator; 2.3 Graphics Processing Units; 2.3.1 NVIDIA's GPU Architecture Overview; 2.3.2 NVIDIA's GPU Architectures; 2.3.3 CUDA Architecture; 2.4 Conclusions; References

3 State-of-the-Art in Pattern Recognition Techniques3.1 Middle Curve Piecewise Linear Approximation; 3.2 Perceptually Important Points; 3.3 Turning Points; 3.4 Symbolic Aggregate approXimation; 3.5 Shapelets; 3.6 Conclusions; References; 4 SAX/GA CPU Approach; 4.1 SAX/GA CPU Approach; 4.1.1 Population Generation; 4.1.2 Fitness Evaluation; 4.1.3 Population Selection; 4.1.4 Chromosome Crossover; 4.1.5 Individual Mutation; 4.2 SAX/GA Performance Analysis; 4.3 Conclusions; References; 5 GPU-Accelerated SAX/GA; 5.1 Parallel SAX Representation; 5.1.1 Prototype 1: SAX Transformation On-Demand

5.1.2 Prototype 2: Speculative FSM5.1.3 Solution A: SAX/GA with Speculative GPU SAX Transformation; 5.2 Parallel Dataset Training; 5.2.1 Prototype 3: Parallel SAX/GA Training; 5.2.2 Solution B: Parallel SAX/GA Training with GPU Fitness Evaluation; 5.3 Fully GPU-Accelerated SAX/GA; 5.3.1 Population Generation Kernel; 5.3.2 Population Selection; 5.3.3 Gene Crossover Kernel; 5.3.4 Gene Mutation Kernel; 5.3.5 Execution Flow; 5.4 Conclusions; Reference; 6 Results; 6.1 SAX/GA Initial Constraints; 6.2 Study Case A: Execution Time; 6.2.1 Solution A: SAX/GA with Speculative FSM

6.2.2 Solution B: Parallel Dataset Training6.2.3 Solution C: Fully GPU-Accelerated SAX/GA; 6.3 Study Case B: FSM Prediction Rate; 6.4 Study Case C: Quality of Solutions; 6.5 Conclusions; 7 Conclusions and Future Work; 7.1 Future Work