Modeling Efficient Classification as a Process of Confidence Assessment and Delegation

In visual object detection and recognition, classifiers have two interesting characteristics: accuracy and speed. Accuracy depends on the complexity of the image features and classifier decision surfaces. Speed depends on the hardware and the computational effort required to use the features and decision surfaces. When attempts to increase accuracy lead to increases in complexity and effort, it is necessary to ask how much are we willing to pay for increased accuracy. For example, if increased computational effort implies quickly diminishing returns in accuracy, then those designing inexpensive surveillance applications cannot aim for maximum accuracy at any cost. It becomes necessary to find trade-offs between accuracy and effort. We study efficient classification of images depicting real-world objects and scenes. Classification is efficient when a classifier can be controlled so that the desired trade-off between accuracy and effort (speed) is achieved and unnecessary computations are avoided on a per input basis. A framework is proposed for understanding and modeling efficient classification of images. Classification is modeled as a tree-like process. In designing the framework, it is important to recognize what is essential and to avoid structures that are narrow in applicability. Earlier frameworks are lacking in this regard. The overall contribution is two-fold. First, the framework is presented, subjected to experiments, and shown to be satisfactory. Second, certain unconventional approaches are experimented with. This allows the separation of the essential from the conventional. To determine if the framework is satisfactory, three categories of questions are identified: trade-off optimization, classifier tree organization, and rules for delegation and confidence modeling. Questions and problems related to each category are addressed and empirical results are presented. For example, related to trade-off optimization, we address the problem of computational bottlenecks that limit the range of trade-offs. We also ask if accuracy versus effort trade-offs can be controlled after training. For another example, regarding classifier tree organization, we first consider the task of organizing a tree in a problem-specific manner. We then ask if problem-specific organization is necessary.

Combat sports analytics: Boxing punch classification using overhead depth imagery

In competitive combat sporting environments like boxing, the statistics on a boxer's performance, including the amount and type of punches thrown, provide a valuable source of data and feedback which is routinely used for coaching and performance improvement purposes. This paper presents a robust framework for the automatic classification of a boxer's punches. Overhead depth imagery is employed to alleviate challenges associated with occlusions, and robust body-part tracking is developed for the noisy time-of-flight sensors. Punch recognition is addressed through both a multi-class SVM and Random Forest classifiers. A coarse-to-fine hierarchical SVM classifier is presented based on prior knowledge of boxing punches. This framework has been applied to shadow boxing image sequences taken at the Australian Institute of Sport with 8 elite boxers. Results demonstrate the effectiveness of the proposed approach, with the hierarchical SVM classifier yielding a 96% accuracy, signifying its suitability for analysing athletes punches in boxing bouts.

Robust and High-resolution Voiced/Unvoiced Classification in Noisy Speech Using A Signal Smoothness Criterion

We propose a novel technique for robust voiced/unvoiced segment detection in noisy speech, based on local polynomial regression. The local polynomial model is well-suited for voiced segments in speech. The unvoiced segments are noise-like and do not exhibit any smooth structure. This property of smoothness is used for devising a new metric called the variance ratio metric, which, after thresholding, indicates the voiced/unvoiced boundaries with 75% accuracy for 0dB global signal-to-noise ratio (SNR). A novelty of our algorithm is that it processes the signal continuously, sample-by-sample rather than frame-by-frame. Simulation results on TIMIT speech database (downsampled to 8kHz) for various SNRs are presented to illustrate the performance of the new algorithm. Results indicate that the algorithm is robust even in high noise levels.

Meltdown - a tool for classification and analysis of DSF data

An application that translates raw thermal melt curve data into more easily assimilated knowledge is described. This program, called ‘Meltdown’, performs a number of data remediation steps before classifying melt curves and estimating melting temperatures. The final output is a report that summarizes the results of a differential scanning fluorimetry experiment. Meltdown uses a Bayesian classification scheme, enabling reproducible identification of various trends commonly found in DSF datasets. The goal of Meltdown is not to replace human analysis of the raw data, but to provide a sensible interpretation of the data to make this useful experimental technique accessible to naïve users, as well as providing a starting point for detailed analyses by more experienced users.

Feature extraction for fingerprint classification

This paper presents two algorithms for smoothing and feature extraction for fingerprint classification. Deutsch's(2) Thinning algorithm (rectangular array) is used for thinning the digitized fingerprint (binary version). A simple algorithm is also suggested for classifying the fingerprints. Experimental results obtained using such algorithms are presented.

Classification of Nonenzymatic Homologues of Protein Kinases

Protein Kinase-Like Non-kinases (PKLNKs), which are closely related to protein kinases, lack the crucial catalytic aspartate in the catalytic loop, and hence cannot function as protein kinase, have been analysed. Using various sensitive sequence analysis methods, we have recognized 82 PKLNKs from four higher eukaryotic organisms, namely, Homo sapiens, Mus musculus, Rattus norvegicus, and Drosophila melanogaster. On the basis of their domain combination and function, PKLNKs have been classified mainly into four categories: (1) Ligand binding PKLNKs, (2) PKLNKs with extracellular protein-protein interaction domain, (3) PKLNKs involved in dimerization, and (4) PKLNKs with cytoplasmic protein-protein interaction module. While members of the first two classes of PKLNKs have transmembrane domain tethered to the PKLNK domain, members of the other two classes of PKLNKs are cytoplasmic in nature. The current classification scheme hopes to provide a convenient framework to classify the PKLNKs from other eukaryotes which would be helpful in deciphering their roles in cellular processes.

Vocalizations of wild Asian elephants (Elephas maximus): Structural classification and social context

Elephants use vocalizations for both long and short distance communication. Whereas the acoustic repertoire of the African elephant (Loxodonta africana) has been extensively studied in its savannah habitat, very little is known about the structure and social context of the vocalizations of the Asian elephant (Elephas maximus), which is mostly found in forests. In this study, the vocal repertoire of wild Asian elephants in southern India was examined. The calls could be classified into four mutually exclusive categories, namely, trumpets, chirps, roars, and rumbles, based on quantitative analyses of their spectral and temporal features. One of the call types, the rumble, exhibited high structural diversity, particularly in the direction and extent of frequency modulation of calls. Juveniles produced three of the four call types, including trumpets, roars, and rumbles, in the context of play and distress. Adults produced trumpets and roars in the context of disturbance, aggression, and play. Chirps were typically produced in situations of confusion and alarm. Rumbles were used for contact calling within and among herds, by matriarchs to assemble the herd, in close-range social interactions, and during disturbance and aggression. Spectral and temporal features of the four call types were similar between Asian and African elephants.

Pi-turns in proteins and peptides: Classification, conformation, occurrence, hydration and sequence.

The i + 5-->i hydrogen bonded turn conformation (pi-turn) with the fifth residue adopting alpha L conformation is frequently found at the C-terminus of helices in proteins and hence is speculated to be a "helix termination signal." An analysis of the occurrence of i + 5-->i hydrogen bonded turn conformation at any general position in proteins (not specifically at the helix C-terminus), using coordinates of 228 protein crystal structures determined by X-ray crystallography to better than 2.5 A resolution is reported in this paper. Of 486 detected pi-turn conformations, 367 have the (i + 4)th residue in alpha L conformation, generally occurring at the C-terminus of alpha-helices, consistent with previous observations. However, a significant number (111) of pi-turn conformations occur with (i + 4)th residue in alpha R conformation also, generally occurring in alpha-helices as distortions either at the terminii or at the middle, a novel finding. These two sets of pi-turn conformations are referred to by the names pi alpha L and pi alpha R-turns, respectively, depending upon whether the (i + 4)th residue adopts alpha L or alpha R conformations. Four pi-turns, named pi alpha L'-turns, were noticed to be mirror images of pi alpha L-turns, and four more pi-turns, which have the (i + 4)th residue in beta conformation and denoted as pi beta-turns, occur as a part of hairpin bend connecting twisted beta-strands. Consecutive pi-turns occur, but only with pi alpha R-turns. The preference for amino acid residues is different in pi alpha L and pi alpha R-turns. However, both show a preference for Pro after the C-termini. Hydrophilic residues are preferred at positions i + 1, i + 2, and i + 3 of pi alpha L-turns, whereas positions i and i + 5 prefer hydrophobic residues. Residue i + 4 in pi alpha L-turns is mainly Gly and less often Asn. Although pi alpha R-turns generally occur as distortions in helices, their amino acid preference is different from that of helices. Poor helix formers, such as His, Tyr, and Asn, also were found to be preferred for pi alpha R-turns, whereas good helix former Ala is not preferred. pi-Turns in peptides provide a picture of the pi-turn at atomic resolution. Only nine peptide-based pi-turns are reported so far, and all of them belong to pi alpha L-turn type with an achiral residue in position i + 4. The results are of importance for structure prediction, modeling, and de novo design of proteins.

Pheo-type: A diagnostic gene-expression assay for the classification of pheochromocytoma and paraganglioma

Context: Pheochromocytomas and paragangliomas (PPGLs) are heritable neoplasms that can be classified into gene-expression subtypes corresponding to their underlying specific genetic drivers. Objective: This study aimed to develop a diagnostic and research tool (Pheo-type) capable of classifying PPGL tumors into gene-expression subtypes that could be used to guide and interpret genetic testing, determine surveillance programs, and aid in elucidation of PPGL biology. Design: A compendium of published microarray data representing 205 PPGL tumors was used for the selection of subtype-specific genes that were then translated to the Nanostring gene-expression platform. A support vector machine was trained on the microarray dataset and then tested on an independent Nanostring dataset representing 38 familial and sporadic cases of PPGL of known genotype (RET, NF1, TMEM127, MAX, HRAS, VHL, and SDHx). Different classifier models involving between three and six subtypes were compared for their discrimination potential. Results: A gene set of 46 genes and six endogenous controls was selected representing six known PPGL subtypes; RTK1–3 (RET, NF1, TMEM127, and HRAS), MAX-like, VHL, and SDHx. Of 38 test cases, 34 (90%) were correctly predicted to six subtypes based on the known genotype to gene-expression subtype association. Removal of the RTK2 subtype from training, characterized by an admixture of tumor and normal adrenal cortex, improved the classification accuracy (35/38). Consolidation of RTK and pseudohypoxic PPGL subtypes to four- and then three-class architectures improved the classification accuracy for clinical application. Conclusions: The Pheo-type gene-expression assay is a reliable method for predicting PPGL genotype using routine diagnostic tumor samples.

Distribution of non-biting midges (Diptera, Chironomidae) in subarctic lakes of Finnish Lapland applications in lake classification and palaeolimnology

Climate change contributes directly or indirectly to changes in species distributions, and there is very high confidence that recent climate warming is already affecting ecosystems. The Arctic has already experienced the greatest regional warming in recent decades, and the trend is continuing. However, studies on the northern ecosystems are scarce compared to more southerly regions. Better understanding of the past and present environmental change is needed to be able to forecast the future. Multivariate methods were used to explore the distributional patterns of chironomids in 50 shallow (≤ 10m) lakes in relation to 24 variables determined in northern Fennoscandia at the ecotonal area from the boreal forest in the south to the orohemiarctic zone in the north. Highest taxon richness was noted at middle elevations around 400 m a.s.l. Significantly lower values were observed from cold lakes situated in the tundra zone. Lake water alkalinity had the strongest positive correlation with the taxon richness. Many taxa had preference for lakes either on tundra area or forested area. The variation in the chironomid abundance data was best correlated with sediment organic content (LOI), lake water total organic carbon content, pH and air temperature, with LOI being the strongest variable. Three major lake groups were separated on the basis of their chironomid assemblages: (i) small and shallow organic-rich lakes, (ii) large and base-rich lakes, and (iii) cold and clear oligotrophic tundra lakes. Environmental variables best discriminating the lake groups were LOI, taxon richness, and Mg. When repeated, this kind of an approach could be useful and efficient in monitoring the effects of global change on species ranges. Many species of fast spreading insects, including chironomids, show a remarkable ability to track environmental changes. Based on this ability, past environmental conditions have been reconstructed using their chitinous remains in the lake sediment profiles. In order to study the Holocene environmental history of subarctic aquatic systems, and quantitatively reconstruct the past temperatures at or near the treeline, long sediment cores covering the last 10000 years (the Holocene) were collected from three lakes. Lower temperature values than expected based on the presence of pine in the catchment during the mid-Holocene were reconstructed from a lake with great water volume and depth. The lake provided thermal refuge for profundal, cold adapted taxa during the warm period. In a shallow lake, the decrease in the reconstructed temperatures during the late Holocene may reflect the indirect response of the midges to climate change through, e.g., pH change. The results from three lakes indicated that the response of chironomids to climate have been more or less indirect. However, concurrent shifts in assemblages of chironomids and vegetation in two lakes during the Holocene time period indicated that the midges together with the terrestrial vegetation had responded to the same ultimate cause, which most likely was the Holocene climate change. This was also supported by the similarity in the long-term trends in faunal succession for the chironomid assemblages in several lakes in the area. In northern Finnish Lapland the distribution of chironomids were significantly correlated with physical and limnological factors that are most likely to change as a result of future climate change. The indirect and individualistic response of aquatic systems, as reconstructed using the chironomid assemblages, to the climate change in the past suggests that in the future, the lake ecosystems in the north do not respond in one predictable way to the global climate change. Lakes in the north may respond to global climate change in various ways that are dependent on the initial characters of the catchment area and the lake.

Two-level k-means clustering algorithm for k-tau relationship establishment and linear-time classification

Partitional clustering algorithms, which partition the dataset into a pre-defined number of clusters, can be broadly classified into two types: algorithms which explicitly take the number of clusters as input and algorithms that take the expected size of a cluster as input. In this paper, we propose a variant of the k-means algorithm and prove that it is more efficient than standard k-means algorithms. An important contribution of this paper is the establishment of a relation between the number of clusters and the size of the clusters in a dataset through the analysis of our algorithm. We also demonstrate that the integration of this algorithm as a pre-processing step in classification algorithms reduces their running-time complexity.

Using multi-label classification for acoustic pattern detection and assisting bird species surveys

Acoustics is a rich source of environmental information that can reflect the ecological dynamics. To deal with the escalating acoustic data, a variety of automated classification techniques have been used for acoustic patterns or scene recognition, including urban soundscapes such as streets and restaurants; and natural soundscapes such as raining and thundering. It is common to classify acoustic patterns under the assumption that a single type of soundscapes present in an audio clip. This assumption is reasonable for some carefully selected audios. However, only few experiments have been focused on classifying simultaneous acoustic patterns in long-duration recordings. This paper proposes a binary relevance based multi-label classification approach to recognise simultaneous acoustic patterns in one-minute audio clips. By utilising acoustic indices as global features and multilayer perceptron as a base classifier, we achieve good classification performance on in-the-field data. Compared with single-label classification, multi-label classification approach provides more detailed information about the distributions of various acoustic patterns in long-duration recordings. These results will merit further biodiversity investigations, such as bird species surveys.

Classification of Indian meteorological stations using cluster and fuzzy cluster analysis, and Kohonen artificial neural networks

The present study deals with the application of cluster analysis, Fuzzy Cluster Analysis (FCA) and Kohonen Artificial Neural Networks (KANN) methods for classification of 159 meteorological stations in India into meteorologically homogeneous groups. Eight parameters, namely latitude, longitude, elevation, average temperature, humidity, wind speed, sunshine hours and solar radiation, are considered as the classification criteria for grouping. The optimal number of groups is determined as 14 based on the Davies-Bouldin index approach. It is observed that the FCA approach performed better than the other two methodologies for the present study.