Dimensionality Reduction Visualization Research Articles

Shape Pattern Recognition of Building Footprints Using t-SNE Dimensionality Reduction Visualization

The shape pattern recognition of building footprints stands as a pivotal concern within GIS spatial cognition. In this study, we introduce a novel approach for the shape recognition of building footprints, leveraging t-distributed stochastic neighbor embedding (t-SNE) dimensionality reduction visualization. First, the Canonical Time Warping (CTW) algorithm is employed to gauge the shape similarity distance of building footprints. Subsequently, the t-SNE model is utilized to map the building footprints, featuring varying numbers of coordinate vertices, onto points within the Cartesian coordinate system. The shape similarity distance serves as the input to the t-SNE model for parameter optimization. Lastly, building footprint shapes are identified through the inherent clustering patterns of points using a Gaussian Mixture Model (GMM). Experimental results demonstrate the method’s robustness to the translation, rotation, scaling, and mirroring of geometric objects, while effectively measuring shape similarity between building footprints. Furthermore, diverse types of building footprints are discernible through natural clustering in low-dimensional spaces, aligning closely with human visual perception.

Single-cell sequencing of the substantia nigra reveals microglial activation in a model of MPTP.

N-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) is a neurotoxin widely used to induce PD models, but the effect of MPTP on the cells and genes of PD has not been fully elucidated. Single-nucleus RNA sequencing was performed in the Substantia Nigra (SN) of MPTP mice. UMAP analysis was used for the dimensionality reduction visualization of the SN in the MPTP mice. Known marker genes highly expressed genes in each cluster were used to annotate most clusters. Specific Differentially Expressed Genes (DEGs) and PD risk genes analysis were used to find MPTP-associated cells. GO, KEGG, PPI network, GSEA and CellChat analysis were used to reveal cell type-specific functional alterations and disruption of cell-cell communication networks. Subset reconstruction and pseudotime analysis were used to reveal the activation status of the cells, and to find the transcription factors with trajectory characterized. Initially, we observed specific DEGs and PD risk genes enrichment in microglia. Next, We obtained the functional phenotype changes in microglia and found that IGF, AGRN and PTN pathways were reduced in MPTP mice. Finally, we analyzed the activation state of microglia and revealed a pro-inflammatory trajectory characterized by transcription factors Nfe2l2 and Runx1. Our work revealed alterations in microglia function, signaling pathways and key genes in the SN of MPTP mice.

Abstract 4956: A fast and efficient bioinformatics analysis workflow for processing reads from single-cell multiomics assays captured on a microwell-based platform

Abstract Here, we present a significant update: version 2.0 of our sequencing analysis, it is a comprehensive primary analysis workflow that is up to 7X faster and consumes 2X less disk space than the previous release. We used it to process a library with 20,000 human PBMCs sequenced to a depth of 1.2 billion reads in 2.5 hours of wall-clock time. The pipeline can be used with reads from single-cell whole transcriptome, targeted mRNA, surface antigens (AbSeq), TCR/BCR and Sample Tag libraries captured on the BD Rhapsody™ System platform. The major processing steps include Quality Filtering, Cell Barcode Identification, Read Alignment, Feature Assignment, UMI Error Correction, Identification of Putative Cells, Sample De-Multiplexing, Dimensionality Reduction for Visualization and VDJ Contig Assembly and Annotation. Output files and metrics are available in easy-to-digest formats, including an html report with dynamic visualization. Integrated outputs in Seurat and Scanpy formats combine expression matrices and all cell annotation metadata (e.g., predicted cell type, sample assignment, TCR/BCR sequence and gene segments). These pre-generated files are ready to load into popular single-cell analysis tools. The pipeline uses CWL as the workflow manager and makes use of custom code written in C++ and Python along with various open-source packages for data processing. The pipeline can also process reads from non-human species and includes built-in support for specifying and building custom reference genome indices. In support of high-throughput multiomic discovery studies enabled by the BD Rhapsody™ HT Xpress System, this pipeline has been tested with datasets containing more than 800,000 putative cells and 14.5 billion reads. For Research Use Only. Not for use in diagnostic or therapeutic procedures. BD, the BD Logo and BD Rhapsody are trademarks of Becton, Dickinson and Company or its affiliates. © 2023 BD. All rights reserved. NPM-2535 (v1.0) 1023 Citation Format: Raghavendra Padmanabhan, Brent Weichel, Amie Radenbaugh, Yuefu Jiang, Charles Weeks, Thomas McCarthy, Youngsook Kim, Anthony Berno, Devon Jensen. A fast and efficient bioinformatics analysis workflow for processing reads from single-cell multiomics assays captured on a microwell-based platform [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4956.

Residual Convolutional Attention Model With Transfer Learning for Detecting Multianomalous Features in Structural Vibration Data

In response to the data anomalies and frequent false alarms caused by harsh environments in long‐term structural health monitoring (SHM), this study has reframed the detection of abnormal vibration data as a time series classification problem. This approach identifies multiple anomalous features, thereby reducing manual detection costs. The novel developed Convolutional Neural Network with Squeeze‐and‐Excitation and Multi‐Head Self‐Attention (CNN–SE–MHSA) employs a deep residual network structure with channel and spatial attention mechanisms, effectively handling the global long‐term dependencies required for anomaly feature learning. It better understands and utilizes feature information across different levels and dimensions, enhancing classification accuracy in complex anomaly situations. Through t‐SNE dimensionality reduction visualization and interpretability analysis, it is demonstrated that the model excels in identifying critical features. Furthermore, by generating simulated data with a variational autoencoder (VAE) and implementing transfer learning strategies based on these data, the issue of low recognition accuracy for complex anomaly data due to data imbalance can be effectively mitigated. In a 25‐day long‐term monitoring experiment of indoor tunnel lining structures, this method demonstrated an average accuracy rate exceeding 96% and a rapid detection capability within 16 min. The results indicate that this method achieves high accuracy in anomaly detection for long‐term monitoring data, even when relying exclusively on time‐domain data.

A 3D Chromosome Structure Reconstruction method with High Resolution Hi-C Data using Nonlinear Dimensionality Reduction and Divide-and-conquer strategy.

Chromosomes are fundamental components of genetic material, and their structural characteristics play an essential role in the regulation of gene expression. The advent of high-resolution Hi-C data has enabled scientists to explore the three-dimensional structure of chromosomes. However, most of the currently available methods for reconstructing chromosome structures are unable to achieve high resolutions, such as 5 Kilobase (KB). In this study, we present NeRV-3D, an innovative method that utilizes a nonlinear dimensionality reduction visualization algorithm to reconstruct 3D chromosome structures at low resolutions. Additionally, we introduce NeRV-3D-DC, which employs a divide-and-conquer technique to reconstruct and visualize 3D chromosome structures at high resolutions. Our results demonstrate that both NeRV-3D and NeRV-3D-DC outperform existing methods in terms of 3D visualization effects and evaluation metrics on simulated and actual Hi-C datasets. The implementation of NeRV-3D-DC can be found at https://github.com/ghaiyan/NeRV-3D-DC.

Dimensionality Reduction and Visualization in Public Management Research from the Perspective of Algorithm Recommendation

With the development of computer technology, algorithm recommendation systems have been widely used in intelligent transportation, urban planning, and other fields. Aiming at the current problems in public management in China, this paper proposes a dimensionality reduction calculation method, and uses the improved particle swarm optimization algorithm to analyze and evaluate the dimensionality reduction process. With the expectation of providing some reference value and ideas for solving practical application problems, a corresponding relationship equation is established with the minimum expected value as the objective function to solve the corresponding optimal solution, the evaluation results of system performance indicators and comprehensive scoring criteria under relevant parameters and constraints are given. The test results show that the visualization effect of the dimensionality reduction visualization model for public management research based on the algorithm recommendation perspective is over 90%.

New fusion features convolutional neural network with high generalization ability on rolling bearing fault diagnosis

SummaryThis study proposes a feature fusion one‐dimensional convolutional neural network algorithm model with a portable non‐dimensionality reduction attention mechanism to resolve the limitations pertaining to the small training sample capacity of conventional deep learning models, single feature‐extractor size, insufficient feature extraction of the bearing faults, and low recognition rate of the bearing health status under variable loads and strong noise. This model regarded raw vibration signal as input, used the feature fusion module to extract multiscale time information, and the convolution‐pooling alternating layer adaptively overcome the time‐dependent limitations. An adaptive batch normalization was introduced to reduce the divergence in sample distribution between the source and target domains and enhance the model generalization capabilities. Moreover, the non‐dimensionality reduction attention mechanism is a more efficient channel weight distribution algorithm that can enhance the anti‐noise performance of the model. After embedding the non‐dimensionality reduction attention mechanism, the proposed network could enhance the anti‐noise performance of the model through a special channel‐weight learning method. It combined the Softmax classification layer to construct a feature extraction feature classification dual intelligent fault diagnosis algorithm. The experiments were performed on the rolling bearing fault dataset. The results revealed that the method provided a rather high generalization ability and stable anti‐noise ability. Simultaneously, the introduction of t‐SNE dimensionality reduction visualization revealed the strong feature‐extraction ability of the deep network model for large‐volume samples.

ScGMAAE: Gaussian mixture adversarial autoencoders for diversification analysis of scRNA-seq data.

The progress of single-cell RNA sequencing (scRNA-seq) has led to a large number of scRNA-seq data, which are widely used in biomedical research. The noise in the raw data and tens of thousands of genes pose a challenge to capture the real structure and effective information of scRNA-seq data. Most of the existing single-cell analysis methods assume that the low-dimensional embedding of the raw data belongs to a Gaussian distribution or a low-dimensional nonlinear space without any prior information, which limits the flexibility and controllability of the model to a great extent. In addition, many existing methods need high computational cost, which makes them difficult to be used to deal with large-scale datasets. Here, we design and develop a depth generation model named Gaussian mixture adversarial autoencoders (scGMAAE), assuming that the low-dimensional embedding of different types of cells follows different Gaussian distributions, integrating Bayesian variational inference and adversarial training, as to give the interpretable latent representation of complex data and discover the statistical distribution of different types of cells. The scGMAAE is provided with good controllability, interpretability and scalability. Therefore, it can process large-scale datasets in a short time and give competitive results. scGMAAE outperforms existing methods in several ways, including dimensionality reduction visualization, cell clustering, differential expression analysis and batch effect removal. Importantly, compared with most deep learning methods, scGMAAE requires less iterations to generate the best results.

HI-MViT: A lightweight model for explainable skin disease classification based on modified MobileViT.

To develop an explainable lightweight skin disease high-precision classification model that can be deployed to the mobile terminal. In this study, we present HI-MViT, a lightweight network for explainable skin disease classification based on Modified MobileViT. HI-MViT is mainly composed of ordinary convolution, Improved-MV2, MobileViT block, global pooling, and fully connected layers. Improved-MV2 uses the combination of shortcut and depth classifiable convolution to substantially decrease the amount of computation while ensuring the efficient implementation of information interaction and memory. The MobileViT block can efficiently encode local and global information. In addition, semantic feature dimensionality reduction visualization and class activation mapping visualization methods are used for HI-MViT to further understand the attention area of the model when learning skin lesion images. The International Skin Imaging Collaboration has assembled and made available the ISIC series dataset. Experiments using the HI-MViT model on the ISIC-2018 dataset achieved scores of 0.931, 0.932, 0.961, and 0.977 on F1-Score, Accuracy, Average Precision (AP), and area under the curve (AUC). Compared with the top five algorithms of ISIC-2018 Task 3, Marco's average F1-Score, AP, and AUC have increased by 6.9%, 6.8%, and 0.8% compared with the suboptimal performance model. Compared with ConvNeXt, the most competitive convolutional neural network architecture, our model is 5.0%, 3.4%, 2.3%, and 2.2% higher in F1-Score, Accuracy, AP, and AUC, respectively. The experiments on the ISIC-2017 dataset also achieved excellent results, and all indicators were better than the top five algorithms of ISIC-2017 Task 3. Using the trained model to test on the PH2 dataset, an excellent performance score is obtained, which shows that it has good generalization performance. The skin disease classification model HI-MViT proposed in this article shows excellent classification performance and generalization performance in experiments. It demonstrates how the classification outcomes can be applied to dermatologists' computer-assisted diagnostics, enabling medical professionals to classify various dermoscopic images more rapidly and reliably.

Integrating Constraints Into Dimensionality Reduction for Visualization: A Survey

This survey reviews and organizes existing methods for integrating constraints into dimensionality reduction (DR). In the world of high-dimensional data, DR methods help to reduce dimensionality while preserving important structures to facilitate subsequent tasks, such as data visualization. While DR methods only reveal hidden structures from the original data, additional information, such as class labels, external features, or even feedback or prior knowledge from users can help to enrich low-dimensional representations. We consider all these types of additional information as constraints. Integrating constraints into classification and clustering methods is well studied, yet, a systematic review on constraint integration in DR methods for visualization is still lacking. We contribute to the literature of constraints in DR visualizations with a novel categorization focusing on constraint types. This survey also introduces new perspectives on the subject, and suggests new trends and future research directions for combining constraints and DR methods.

Development of a deep learning algorithm for myopic maculopathy classification based on OCT images using transfer learning.

PurposeTo apply deep learning (DL) techniques to develop an automatic intelligent classification system identifying the specific types of myopic maculopathy (MM) based on macular optical coherence tomography (OCT) images using transfer learning (TL).MethodIn this retrospective study, a total of 3,945 macular OCT images from 2,866 myopic patients were recruited from the ophthalmic outpatients of three hospitals. After culling out 545 images with poor quality, a dataset containing 3,400 macular OCT images was manually classified according to the ATN system, containing four types of MM with high OCT diagnostic values. Two DL classification algorithms were trained to identify the targeted lesion categories: Algorithm A was trained from scratch, and algorithm B using the TL approach initiated from the classification algorithm developed in our previous study. After comparing the training process, the algorithm with better performance was tested and validated. The performance of the classification algorithm in the test and validation sets was evaluated using metrics including sensitivity, specificity, accuracy, quadratic-weighted kappa score, and the area under the receiver operating characteristic curve (AUC). Moreover, the human-machine comparison was conducted. To better evaluate the algorithm and clarify the optimization direction, the dimensionality reduction analysis and heat map analysis were also used to visually analyze the algorithm.ResultsAlgorithm B showed better performance in the training process. In the test set, the algorithm B achieved relatively robust performance with macro AUC, accuracy, and quadratic-weighted kappa of 0.986, 96.04% (95% CI: 0.951, 0.969), and 0.940 (95% CI: 0.909–0.971), respectively. In the external validation set, the performance of algorithm B was slightly inferior to that in the test set. In human-machine comparison test, the algorithm indicators were inferior to the retinal specialists but were the same as the ordinary ophthalmologists. In addition, dimensionality reduction visualization and heatmap visualization analysis showed excellent performance of the algorithm.ConclusionOur macular OCT image classification algorithm developed using the TL approach exhibited excellent performance. The automatic diagnosis system for macular OCT images of MM based on DL showed potential application prospects.

Network representation learning method embedding linear and nonlinear network structures

With the rapid development of neural networks, much attention has been focused on network embedding for complex network data, which aims to learn low-dimensional embedding of nodes in the network and how to effectively apply learned network representations to various graph-based analytical tasks. Two typical models exist namely the shallow random walk network representation method and deep learning models such as graph convolution networks (GCNs). The former one can be used to capture the linear structure of the network using depth-first search (DFS) and width-first search (BFS), whereas Hierarchical GCN (HGCN) is an unsupervised graph embedding that can be used to describe the global nonlinear structure of the network via aggregating node information. However, the two existing kinds of models cannot simultaneously capture the nonlinear and linear structure information of nodes. Thus, the nodal characteristics of nonlinear and linear structures are explored in this paper, and an unsupervised representation method based on HGCN that joins learning of shallow and deep models is proposed. Experiments on node classification and dimension reduction visualization are carried out on citation, language, and traffic networks. The results show that, compared with the existing shallow network representation model and deep network model, the proposed model achieves better performances in terms of micro-F1, macro-F1 and accuracy scores.

Aircraft Trajectory Clustering in Terminal Airspace Based on Deep Autoencoder and Gaussian Mixture Model

The aircraft trajectory clustering analysis in the terminal airspace is conducive to determining the representative route structure of the arrival and departure trajectory and extracting their typical patterns, which is important for air traffic management such as airspace structure optimization, trajectory planning, and trajectory prediction. However, the current clustering methods perform poorly due to the large flight traffic, high density, and complex airspace structure in the terminal airspace. In recent years, the continuous development of Deep Learning has demonstrated its powerful ability to extract internal potential features of large dataset. Therefore, this paper mainly tries a deep trajectory clustering method based on deep autoencoder (DAE). To this end, this paper proposes a trajectory clustering method based on deep autoencoder (DAE) and Gaussian mixture model (GMM) to mine the prevailing traffic flow patterns in the terminal airspace. The DAE is trained to extract feature representations from historical high-dimensional trajectory data. Subsequently, the output of DAE is input into GMM for clustering. This paper takes the terminal airspace of Guangzhou Baiyun International Airport in China as a case to verify the proposed method. Through the direct visualization and dimensionality reduction visualization of the clustering results, it is found that the traffic flow patterns identified by the clustering method in this paper are intuitive and separable.

IXVC: An interactive pipeline for explaining visual clusters in dimensionality reduction visualizations with decision trees

High-dimensional data with many features are usually challenging to represent with standard visualization techniques. Usually, one has to resort to dimensionality reduction techniques such as PCA, MDS or t -SNE to represent such data. Such dimensionality reduction techniques make it possible to highlight the high-dimensional structures of data. In many of such visualizations, comparable instances appear to form visual clusters. However, no feedback is directly given by these techniques to the user about the features that make the instances cluster together in the visualization. As such, the interpretation of which features define a given visual cluster is a complicated task. In this paper, we propose a novel interactive approach (called Interactive eXplanation of Visual Clusters — IXVC) to explain dimensionality reduction visualizations by mapping their clusters to explanations provided by decision trees. The decision trees use features in high-dimensional data to explain two-dimensional clusters, filling the gap between the dimensionality reduction visualization and the original data.

The incubation effect among students playing an educational game for physics

The incubation effect (IE) is a problem-solving phenomenon composed of three phases: pre-incubation where one fails to solve a problem; incubation, a momentary break where time is spent away from the unsolved problem; and post-incubation where the unsolved problem is revisited and solved. Literature on IE was limited to experiments involving traditional classroom activities. This initial investigation showed evidence of IE instances in a computer-based learning environment. This paper consolidates the studies on IE among students playing an educational game called Physics Playground and presents further analysis to examine the incidence of post-incubation or the revisit to a previously unsolved problem. Prior work, which focused on predicting successful outcomes, includes a coarse-grained IE model developed with logistic regression on aggregated data and an improved model which leveraged long short-term memory (LSTM) combined with dimensionality reduction visualization technique and clustering on fine-grained data. The additional analysis which aims to understand factors that may trigger the post-incubation phase also used fine-grained data and LSTM to create a revisit model. Results show that time elapsed relative to the activity period and encountering a problem with a similar solution during incubation were possible factors in revisiting previously unsolved problems.

Cross-platform comparison of immune-related gene expression to assess intratumor immune responses following cancer immunotherapy

Neoadjuvant immunotherapy can induce immune responses within the tumor microenvironment. Gene expression can be used to assess responses with limited amounts of conventionally-fixed patient-derived samples. We aim to assess the cross-platform concordance of immune-related gene expression data. We performed comparisons across three panels in two platforms: Nanostring nCounter® PanCancer Immune Profiling Panel (nS), HTG EdgeSeq Oncology Biomarker Panel (HTG OBP) and Precision Immuno-Oncology Panel (HTG PIP). All tissue samples of 14 neoadjuvant GM-CSF treated, 14 neoadjuvant Provenge treated, and 12 untreated prostate cancer patients were radical prostatectomy (RP) tissues, while 6 prostatitis patients and 6 non-prostatitis subjects were biopsies. For all 52 patients, more than 90% of the common genes were significantly correlated (p < 0.05) and more than 76% of the common genes were highly correlated (r > 0.5) between any two panels. Co-inertia analysis also demonstrated high overall dataset structure similarity (correlation>0.84). Although both dimensionality reduction visualization analysis and unsupervised hierarchical cluster analysis for highly correlated common genes (r > 0.9) suggested a high-level of consistency across the panels, there were subsets of genes that were differentially expressed across the panels. In addition, while the effect size of the differential testing for neoadjuvant treated vs. untreated localized prostate cancer patients across the panels were significantly correlated, some genes were only differentially expressed in the HTG panels. Finally, the HTG PIP panel had the best classification performance among the 3 panels. These differences detected may be a result of the different panels or platforms due to their technical setting and focus. Thus, researchers should be aware of those potential differences when deciding which platform and panel to use.

Analysis and Visualization of Deep Neural Networks in Device-Free Wi-Fi Indoor Localization

Device-free Wi-Fi indoor localization has received significant attention as a key enabling technology for many Internet of Things (IoT) applications. Machine learning-based location estimators, such as the deep neural network (DNN), carry proven potential in achieving high-precision localization performance by automatically learning discriminative features from the noisy wireless signal measurements. However, the inner workings of DNNs are not transparent and not adequately understood especially in the indoor localization application. In this paper, we provide quantitative and visual explanations for the DNN learning process as well as the critical features that DNN has learned during the process. Toward this end, we propose to use several visualization techniques, including: 1) dimensionality reduction visualization, to project the high-dimensional feature space to the 2D space to facilitate visualization and interpretation, and 2) visual analytics and information visualization, to quantify relative contributions of each feature with the proposed feature manipulation procedures. The results provide insightful views and plausible explanations of the DNN in device-free Wi-Fi indoor localization using channel state information (CSI) fingerprints.

Visualizing nationwide variation in medicare Part D prescribing patterns

BackgroundTo characterize the regional and national variation in prescribing patterns in the Medicare Part D program using dimensional reduction visualization methods.MethodsUsing publicly available Medicare Part D claims data, we identified and visualized regional and national provider prescribing profile variation with unsupervised clustering and t-distributed stochastic neighbor embedding (t-SNE) dimensional reduction techniques. Additionally, we examined differences between regionally representative prescribing patterns for major metropolitan areas.ResultsDistributions of prescribing volume and medication diversity were highly skewed among over 800,000 Medicare Part D providers. Medical specialties had characteristic prescribing patterns. Although the number of Medicare providers in each state was highly correlated with the number of Medicare Part D enrollees, some states were enriched for providers with > 10,000 prescription claims annually. Dimension-reduction, hierarchical clustering and t-SNE visualization of drug- or drug-class prescribing patterns revealed that providers cluster strongly based on specialty and sub-specialty, with large regional variations in prescribing patterns. Major metropolitan areas had distinct prescribing patterns that tended to group by major geographical divisions.ConclusionsThis work demonstrates that unsupervised clustering, dimension-reduction and t-SNE visualization can be used to analyze and visualize variation in provider prescribing patterns on a national level across thousands of medications, revealing substantial prescribing variation both between and within specialties, regionally, and between major metropolitan areas. These methods offer an alternative system-wide and pattern-centric view of such data for hypothesis generation, visualization, and pattern identification.

Data Driven Models for Satellite State-of-Health Monitoring and Evaluation

Satellite state-of-health conditions and keeping up systems safe are the fundamental worry of space operations. Since managing the correlation structure between variables give greater information than individual variables or causes relationships, Data driven methodology presented as an answer of managing enormous information. In this article a survey of some recent related work is done. Then a mathematical foundation for latent variable modeling, support vector machines (SVMs) and artificial neural networks (ANN) is illustrated. Then a four data driven models are introduced for monitoring, diagnosis and prediction of the thermal control of satellite power supply system. Two kinds of PLS-Batch models were implemented for the first time with a satellite telemetry. After that a utilization of multivariate Shewart, DModX in addition to CuSum charts for system monitoring and early fault detection. Another two machine learning models were built and coded in Python. These two models are (SVMs) model and (ANN) model. Both of the last two models are integrated with PCA for feature extraction and dimensionality reduction for visualization. The outcomes confirmed the effectiveness of utilizing such models in monitoring, faults detection, and diagnosis. Also system health evaluation could be established within the operation through these models. In addition to prediction of system behavior and early faults detection.

A Perception-Driven Approach to Supervised Dimensionality Reduction for Visualization.

Dimensionality reduction (DR) is a common strategy for visual analysis of labeled high-dimensional data. Low-dimensional representations of the data help, for instance, to explore the class separability and the spatial distribution of the data. Widely-used unsupervised DR methods like PCA do not aim to maximize the class separation, while supervised DR methods like LDA often assume certain spatial distributions and do not take perceptual capabilities of humans into account. These issues make them ineffective for complicated class structures. Towards filling this gap, we present a perception-driven linear dimensionality reduction approach that maximizes the perceived class separation in projections. Our approach builds on recent developments in perception-based separation measures that have achieved good results in imitating human perception. We extend these measures to be density-aware and incorporate them into a customized simulated annealing algorithm, which can rapidly generate a near optimal DR projection. We demonstrate the effectiveness of our approach by comparing it to state-of-the-art DR methods on 93 datasets, using both quantitative measure and human judgments. We also provide case studies with class-imbalanced and unlabeled data.