Combining wavelength importance ranking to the random forest classifier to analyze multiclass spectral data

Fontes, Juliana de Abreu; Anzanello, Michel J.; Brito, João B.G.; Bucco, Guilherme Brandelli; Fogliatto, Flávio Sanson; Puglia, Fábio do Prado

doi:10.1016/j.forsciint.2021.110998

Cited by 11 publications

(17 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, it was interesting that for our dataset Random Forest performed slightly better than Multilayer Perceptron algorithms when using the Partial feature set (accuracies for best candidates were 88% and 87%, respectively). Two possible factors could be involved: (1) the nature of Random Forest, which is based on conditional decisions, allows better handling of a mixture of categorical and numerical features 42 ; and (2) the relatively small dataset size of our study (total dataset: 385, and training dataset: 308) may be a more pronounced constraint for Multilayer Perceptron, which typically requires large datasets. As an example, using the Partial feature set (16 features), the number of parameters for a relatively simple two-layer MLP (MLP-Partial-10-16) will be 373 (for weights: 16*10 + 10*16 + 16*1 = 336, and for biases: 10 + 16 + 1 = 27), which is greater than the size of our training instances and potentially results in underfitting of the predictive models.…”

Section: Discussionmentioning

confidence: 99%

NeurostimML: A machine learning model for predicting neurostimulation-induced tissue damage

Li,

Frederick,

George

et al. 2023

Preprint

View full text Add to dashboard Cite

Objective. The safe delivery of electrical current to neural tissue depends on many factors, yet previous methods for predicting tissue damage rely on only a few stimulation parameters. Here, we report the development of a machine learning approach that could lead to a more reliable method for predicting electrical stimulation-induced tissue damage by incorporating additional stimulation parameters. Approach. A literature search was conducted to build an initial database of tissue response information after electrical stimulation, categorized as either damaging or non-damaging. Subsequently, we used ordinal encoding and random forest for feature selection, and investigated four machine learning models for classification: Logistic Regression, K-nearest Neighbor, Random Forest, and Multilayer Perceptron. Finally, we compared the results of these models against the accuracy of the Shannon equation. Main Results. We compiled a database with 387 unique stimulation parameter combinations collected from 58 independent studies conducted over a period of 47 years, with 195 (51%) categorized as non-damaging and 190 (49%) categorized as damaging. The features selected for building our model with a Random Forest algorithm were: waveform shape, geometric surface area, pulse width, frequency, pulse amplitude, charge per phase, charge density, current density, duty cycle, daily stimulation duration, daily number of pulses delivered, and daily accumulated charge. The Shannon equation yielded an accuracy of 63.9% using a k value of 1.79. In contrast, the Random Forest algorithm was able to robustly predict whether a set of stimulation parameters was classified as damaging or non-damaging with an accuracy of 88.3%. Significance. This novel Random Forest model can facilitate more informed decision making in the selection of neuromodulation parameters for both research studies and clinical practice. This study represents the first approach to use machine learning in the prediction of stimulation-induced neural tissue damage, and lays the groundwork for neurostimulation driven by machine learning models.

show abstract

Section: Discussionmentioning

confidence: 99%

NeurostimML: A machine learning model for predicting neurostimulation-induced tissue damage

Li,

Frederick,

George

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The third approach integrates RF and GI under the SFFS algorithm (identified as Phase 2 with RF-GI in Table 5), which is equivalent to traditional wrapperbased methods widely described in the literature. 19 The best result for each dataset is highlighted in bold.…”

Section: The Importance Of the Two-phase Approach For Wavelength Sele...mentioning

confidence: 99%

“…Numerous techniques have relied on importance indices for guiding the removal of less informative features. 8,10,[18][19][20][21][22] Another effective yet less employed strategy for selecting the most informative wavelengths in complex datasets is wavelength clustering. In this concept, wavelengths relying on similar information are inserted into clusters, and a supervised algorithm is applied to each cluster separately.…”

mentioning

confidence: 99%

A novel two‐phase near‐infrared and midinfrared wavelength selection framework for sample classification

Fontes,

Anzanello,

Brito

et al. 2024

Journal of Chemometrics

View full text Add to dashboard Cite

Spectral data describing product samples are typically composed of a large number of noisy and irrelevant wavelengths that tends to undermine the performance of multivariate predictive techniques. This paper proposes a two‐phase framework that integrates a preselection wavelength step oriented by wavelength clustering to a wrapper‐based strategy. The first phase performs a pruning process in the data that removes the less informative wavelengths relying on the spectral clustering, a technique deemed suitable to the Fourier transform infrared (FTIR) spectroscopy and near‐infrared (NIR) spectroscopy data at hand. The preselected wavelengths undergo a second phase of selection efforts based on the combination of different wavelength importance indices (i.e., Bhattacharyya distance, Chi‐square, ReliefF, and Gini) and classification techniques (i.e., support vector machine, k‐nearest neighbors, and random forest). When applied to 11 FTIR datasets from different domains, the recommended combination of importance index and classifier increased the average accuracy by 6.37% (from 0.863 to 0.918), while retaining average 3.84% of the original spectra. The framework also improved the selection process regarding computational time.

show abstract

“…For this model, the χ2 is the variable selection method, and RF is the classification technique. 19 As samples become more complex, it can take multiple analysis methods to completely characterize a sample. Iakab et al proposed a multimodal approach for analyzing biological samples that combines the results of mass spectrometry and vibrational spectroscopy for imaging to get spatial resolution and molecular information.…”

Section: ■ Introductionmentioning

confidence: 99%

“…To classify data like illicit substances, they utilized RF to select areas of interest and combine the wavelengths of importance into a model named χ2–RF. For this model, the χ2 is the variable selection method, and RF is the classification technique …”

Section: Introductionmentioning

confidence: 99%

Innovative Vibrational Spectroscopy Research for Forensic Application

et al. 2023

View full text Add to dashboard Cite

Combining wavelength importance ranking to the random forest classifier to analyze multiclass spectral data

Cited by 11 publications

References 38 publications

NeurostimML: A machine learning model for predicting neurostimulation-induced tissue damage

NeurostimML: A machine learning model for predicting neurostimulation-induced tissue damage

A novel two‐phase near‐infrared and midinfrared wavelength selection framework for sample classification

Innovative Vibrational Spectroscopy Research for Forensic Application

Contact Info

Product

Resources

About