Automated classification of software issue reports using machine learning techniques: an empirical study

Software developers, testers and customers routinely submit issue reports to software issue trackers to record the problems they face in using a software. The issues are then directed to appropriate experts for analysis and fixing. However, submitters often misclassify an improvement request as a bug and vice versa. This costs valuable developer time. Hence automated classification of the submitted reports would be of great practical utility. In this paper, we analyze how machine learning techniques may be used to perform this task. We apply different classification algorithms, namely naive Bayes, linear discriminant analysis, k-nearest neighbors, support vector machine (SVM) with various kernels, decision tree and random forest separately to classify the reports from three open-source projects. We evaluate their performance in terms of F-measure, average accuracy and weighted average F-measure. Our experiments show that random forests perform best, while SVM with certain kernels also achieve high performance.     

Molecular Engineering of Photoacoustic Performance by Chalcogenide Variation in Conjugated Polymer Nanoparticles for Brain Vascular Imaging

As conjugated polymer nanoparticles (CPNs) have attracted growing interest as photoacoustic (PA) imaging contrast agents, revelation of the relationship between the molecular structure of conjugated polymers and PA property is highly in demand. Here, three donor–acceptor‐structured conjugated polymer analogs are designed, where only a single heteroatom of acceptor units changes from oxygen to sulfur to selenium, allowing for systematic investigation of the molecular structure–PA property relationship. The absorption and PA spectra of these CPNs can be facilely tuned by changing the heteroatoms of the acceptor units. Moreover, the absorption coefficient, and in turn the PA signal intensity, decreases when the heteroatom changes from oxygen to sulfur to selenium. As these CPNs exhibit weak fluorescence and similar photothermal conversion efficiency (≈70%), their PA intensities are approximately proportional to their absorption coefficients. The in vivo brain vasculature imaging in this study also demonstrates this trend. This study provides a simple but efficient strategy to manipulate the PA properties of CPNs through changing the heteroatom at key positions.     

Decoding individuated finger movements using volume-constrained neuronal ensembles in the M1 hand area.

Individuated finger and wrist movements can be decoded using random subpopulations of neurons that are widely distributed in the primary motor (M1) hand area. This work investigates 1) whether it is possible to decode dexterous finger movements using spatially-constrained volumes of neurons as typically recorded from a microelectrode array; and 2) whether decoding accuracy differs due to the configuration or location of the array within the M1 hand area. Single-unit activities were sequentially recorded from task-related neurons in two rhesus monkeys as they performed individuated movements of the fingers and the wrist. Simultaneous neuronal ensembles were simulated by constraining these activities to the recording field dimensions of conventional microelectrode array architectures. Artificial neural network (ANN) based filters were able to decode individuated finger movements with greater than 90% accuracy for the majority of movement types, using as few as 20 neurons from these ensemble activities. Furthermore, for the large majority of cases there were no significant differences (p < 0.01) in decoding accuracy as a function of the location of the recording volume. The results suggest that a brain-machine interface (BMI) for dexterous control of individuated fingers and the wrist can be implemented using microelectrode arrays placed broadly in the M1 hand area.     

Human ECoG analysis during speech perception using matching pursuit: a comparison between stochastic and dyadic dictionaries

We use the matching pursuit (MP) algorithm to detect induced gamma activity in human EEG during speech perception. We show that the MP algorithm is particularly useful for detecting small power changes at high gamma frequencies (> 70 Hz). We also compare the performance of the MP using a stochastic versus a dyadic dictionary and show that despite the frequency bias the time-frequency power plot (averaged over 100 trials) generated by the dyadic MP is almost identical (> 98.5%) to the one generated by the stochastic MP. However, the dyadic MP is computationally much faster than the stochastic MP.     

Phase-dependent effects of spinal cord stimulation on locomotor activity.

This paper examines how electrical stimulation of the spinal cord can modulate the output of the central pattern generator (CPG) for locomotion. Application of discrete current pulses to a single spinal segment was shown to affect multiple parameters of an ongoing locomotor pattern in an in vitro spinal cord. For any given stimulus, the effects on frequency, duration, and symmetry of locomotor output were strongly dependent on the phase at which stimulation was applied within the CPG cycle. Additionally, most stimuli had an immediate impact and evinced no effects on subsequent cycles. The most dramatic changes were seen when stimulation was applied during motor bursting: stimuli applied to the ipsilateral spinal hemicord increased the burst length, while stimuli applied to the contralateral spinal hemicord decreased the burst length. Smaller changes were observed when stimulating during delays between motor bursts. Thus, phasic stimulation was shown to influence the behavior of the CPG and spinal locomotion circuits on a cycle-by-cycle basis. This work represents the first step toward our ultimate goal of developing a neuroprosthetic device to restore locomotion after a severe spinal cord injury.     

Selective detection of hydrogen sulfide using copper oxide-doped tin oxide based thick film sensor array

In this work, copper oxide-doped (1, 3 and 5 wt%) tin oxide powders have been synthesised by sol–gel method and thick film sensor array has been developed by screen printing technique for the detection of H₂S gas. Powder X-ray diffraction pattern shows that the tin oxide (SnO₂) doped with 3 wt% copper oxide (CuO) has smaller crystallite size in comparison to 0, 1 and 5 wt% CuO-doped SnO₂. Furthermore, field emission scanning electron microscopy manifests the formation of porous film consisting of loosely interconnected small crystallites. The effect of various amounts of CuO dopant has been studied on the sensing properties of sensor array with respect to hydrogen sulfide (H₂S) gas. It is found that the SnO₂ doped with 3 wt% CuO is extremely sensitive (82%) to H₂S gas at 150 °C, while it is almost insensitive to many other gases, i.e., hydrogen (H₂), carbon monoxide (CO), sulphur dioxide (SO₂) and liquefied petroleum gas (LPG). Moreover, at low concentration of gas, it shows fast recovery as compared to response time. Such high performance of 3 wt% CuO-doped SnO₂ thick film sensor is probably due to the diminishing of the p–n junction and the smallest crystallite size (11 nm) along with porous structure.     

Estimation of the (n,m) concentration distribution of single-walled carbon nanotubes from photoabsorption spectra

Deconvolution of the absorption spectrum of single-walled carbon nanotubes (SWNTs) into distinct (n,m) contributions is complicated because transition energies are closely spaced. The algorithm presented in this work attempts to simplify the problem by grouping nanotubes with similar transition energies and assigning weights to their spectral contributions. Voigt line shapes were used to fit absorption spectra of sodium dodecyl sulfate suspended HiPco SWNT and CoMoCat SWNT. Line widths for the metallic (93.42 meV) and two semiconducting regions (57.96 and 29.86 meV) were obtained from the absorption spectra of DNA-wrapped SWNT fractionated by ion-exchange chromatography. The method is used to describe the reaction kinetics of certain HiPco SWNTs upon reaction with 4-chlorobenzene diazonium and 4-hydroxybenzene diazonium salts. The code for deconvolution has been provided as open source in the Supporting Information for future modifications.     

Tailoring the Angular Mismatch in MoS2 Homobilayers through Deformation Fields

Ultrathin MoS₂ has shown remarkable characteristics at the atomic scale with an immutable disorder to weak external stimuli. Ion beam modification unlocks the potential to selectively tune the size, concentration, and morphology of defects produced at the site of impact in 2D materials. Combining experiments, first-principles calculations, atomistic simulations, and transfer learning, it is shown that irradiation-induced defects can induce a rotation-dependent moiré pattern in vertically stacked homobilayers of MoS₂ by deforming the atomically thin material and exciting surface acoustic waves (SAWs). Additionally, the direct correlation between stress and lattice disorder by probing the intrinsic defects and atomic environments are demonstrated. The method introduced in this paper sheds light on how engineering defects in the lattice can be used to tailor the angular mismatch in van der Waals (vdW) solids.     

RA-ECDM of Silicon Wafers Using Taguchi’s Methodology and Machine Learning Algorithms

Silicon - As a prominent machining process, electrochemical discharge machining (ECDM) is used to process materials that are both fragile and difficult to cut. The growing use of this method for...     

A Novel Approach to Investigate the Impact of Hetero-High-K Gate Stack on SiGe Junctionless Gate-All-Around (JL-GAA) MOSFET

Silicon - It is a well-known fact that the gate stacking is used to improve the electrostatic behavior of Si0.5Ge0.5 Junctionless Gate-All-Around (JL-GAA) MOSFETs. In gate stacking, the high-k...