Informatics and Machine Learning. Stephen Winters-Hilt

1.3 Chunking on a dynamic table. Works for a HMM using a simple join ...Figure 1.4 Edge feature enhancement via HMM/EM EVA filter. The filter “proje...Figure 1.5 (Left) The general stochastic sequential analysis flow topology. ...

      2 Chapter 2Figure 2.1 The Norwalk virus genome (the “cruise ship virus”).Figure 2.2 The start of the E. coli genome file, FASTA format.Figure 2.3 The Geometric distribution, P(X = k) = (1 − p)^(k−1) p, with...Figure 2.4 The Gaussian distribution, aka Normal, shown with mean zero and v...

      3 Chapter 3Figure 3.1 Codon structure is revealed in the V. cholera genome by mutual in...Figure 3.2 ORF encoding structure is revealed in the V. cholera genome by ga...Figure 3.3 (a) Topology index histograms shown for the V. cholerae CHR. I ge...Figure 3.4 Topology‐index histograms are shown for the Chlamydia trachomatis...Figure 3.5 Hash interpolated Markov model (hIMM) and gap/hash interpolated M...

      4 Chapter 4Figure 4.1 Schematic for the finite state automaton used for acquisition of ...Figure 4.2 Sensitivity (SN) and Specificity (SP). For the predictor evaluato...Figure 4.3 Sensitivity (SN) and Specificity (SP) for other two conventions (...Figure 4.4 FSA with alternating SP:SN optimized tuning. Step 1: Acquire sign...Figure 4.5 Tuning on “start_drop_value for a collection of blockade signals ...Figure 4.6 Robust Spike feature extraction: radiated DNA. A time‐domain FSA ...Figure 4.7 SVM classification results with and without spike analysis. Addin...Figure 4.8 FSA acquisition flowchart.Figure 4.9 Prokaryotic gene structure discovered thus far.Figure 4.10 Two types of “stop” codon.Figure 4.11 Hypothesized splice signal upstream.Figure 4.12 Hypothesized splice signal downstream of stop upstream from true...

      5 Chapter 6Figure 6.1 The most common stochastic sequential analysis flow topology. The...Figure 6.2 Graphical model for a first‐order Markov model.Figure 6.3 Graphical model for a standard hidden Markov model.Figure 6.4 The most probable state sequence.Figure 6.5 Viterbi path. Optimal path identification.Figure 6.6 Comparison of standard HMM and the clique‐generalized HMM. The up...Figure 6.7 HMM/EM Viterbi‐path level occupation feature extraction. Strong E...Figure 6.8 The binary classification performance using features extracted wi...Figure 6.9 Adaboost feature selection strengthens the SVM performance of the...Figure 6.10 If Adaboost operates from the 150 component manual set, a reduce...Figure 6.11 AdaBoosting to select 100 of the full set of 2600 features impro...

      6 Chapter 7Figure 7.1 Top Panel. Sliding‐window association (clique) of observations an...Figure 7.2 The transition schematic for the HMM with duration (HMMD).Figure 7.3 Viterbi column‐pointer match de‐segmentation rule. Table 1 and Ta...

      7 Chapter 8Figure 8.1 Map.Figure 8.2 Chart.Figure 8.3 Overlapping charts.Figure 8.4 Curve. Note: The parameterization λ defines different curves...Figure 8.5 Function.Figure 8.6 Tangent Bundle, where “cross‐section” of TM (gives intuitive noti...Figure 8.7 1 Forms and cotangent bundle. There is a duality between 1 forms ...Figure 8.8 Gradient 1 form.Figure 8.9 Parallel transport.

      8 Chapter 9Figure 9.1 Single Neuron with step (threshold) activation function. The inpu...Figure 9.2 Perceptron update. The update rule shifts the position of the sep...Figure 9.3 Single Neuron, Sigma activation function: the inputs x_k are multi...Figure 9.4 Neural Net, layered topology, fully connected between layers (“de...Figure 9.5 Link functions for GD (linear), EGU^+/− (left and right curv...Figure 9.6 Link functions for BEG (sigmoid curve), where f(ω) =

, L <

      9 Chapter 10Figure 10.1 Supervised learning: separability and maximum margin. Shown in t...Figure 10.2 Support vectors on margin boundary.Figure 10.3 (Left) One of the former positives (central to the positive clus...Figure 10.4 Separable solution with maximum width margin.Figure 10.5 Post‐training, have decision hyperplane. Now test data (ci...Figure 10.6 Two clusters, roughly, shown, with intra‐cluster distance shown ...Figure 10.7 Schematic for SVM‐based clustering, starting with randomly label...Figure 10.8 Sum of squared error (SSE) scoring. SSE, the total of the sum of...Figure 10.9 Perceptron update. The update rule shifts the position of the se...Figure 10.10 Decision boundary (solid line); with margin (region between dot...Figure 10.11 The solid circle is the decision boundary. In the implementatio...Figure 10.12 A possible modeling configuration for two‐cluster explicit situ...Figure 10.13 The Hoeffding inequality can be reduced to P[∣ν − μ ∣ > ɛ...Figure 10.14 Line Ax + By + C = 0 and a nearby point. D =

Figure 10.15 Hyperplane separability.

.Figure 10.16 IF C > α_i, σ_i → ∅ ; If C = α_i, σ_i free(≥∅)...Figure 10.17 Comparative results are shown on performance of Kernels and alg...Figure 10.18 SVM convergence failure seen with 100% SV passing on distribute...Figure 10.19 Sequential learning topology SV pass‐tuning. Dataset = 9G...Figure 10.20 Distributed learning topology SV pass‐tuning. Dataset = 9...Figure 10.21 SMO (non‐chunking) support vector reduction. Dataset: 9GC...Figure 10.22 Sequential chunking support vector reduction. Dataset: 9GC9CG_9...Figure 10.23 Multi‐threaded chunking support vector reduction. Dataset...Figure 10.24 Decision boundary (solid line); with margin (region between dot...Figure 10.25 Nanopore detector signal analysis architecture, with use of an ...Figure 10.26 The percentage data rejection vs. SN + SP curves are shown for ...Figure 10.27 The percentage data rejection vs. SN + SP curves are shown for ...Figure 10.28 The percent increase in iterations‐to‐convergence against the “Figure 10.29 The number of bounded support vectors (BSV) as a function of “CFigure 10.30 (enlarged version of Figure 10.7) SVM‐external clustering metho...Figure 10.31 Summary of the degradation in clustering performance for less o...Figure 10.32 Efforts to use simulated annealing in the number of KKT violato...Figure 10.33 The results of SVM‐relabeler algorithm using a third degree pol...Figure 10.34 SVM‐external clustering results. (a) and (b) show the boost in ...Figure 10.35 The result of relabeler algorithm with perturbation. The top pl...Figure 10.36 (a, b) represent the SSE and purity evaluation of hybrid Re‐lab...Figure 10.37 Clustering performance comparisons: SVM‐external clustering com...Figure 10.38 Nanopore feature vector data (in standard 150 component, L₁‐nor...Figure 10.39 (a) Simulated annealing with constant perturbation, (b) simulat...Figure 10.40 Multiple‐convergence, SVM‐external clustering. Three multiple c...

      10 Chapter