Tropical altitudinal gradient soil organic carbon and nitrogen estimation using Specim IQ portable imaging spectrometer

The largest actively cycling terrestrial carbon pool, soil, has been disturbed during latest centuries by human actions through reduction of woody land cover. Soil organic carbon (SOC) content can reliably be estimated in laboratory conditions, but more cost-efficient and mobile techniques are needed for large-scale monitoring of SOC e.g. in remote areas. We demonstrate the capability of a mobile hyperspectral camera operating in the visible-near infrared wavelength range for practical estimation of soil organic carbon (SOC) and nitrogen content, to support efficient monitoring of soil properties. The 191 soil samples were collected in Taita Taveta County, Kenya representing an altitudinal gradient comprising five typical land use types: agroforestry, cropland, forest, shrubland and sisal estate. The soil samples were imaged using a Specim IQ hyperspectral camera under controlled laboratory conditions, and their carbon and nitrogen content was determined with a combustion analyzer. We use machine learning for estimating SOC and N content based on the spectral images, studying also automatic selection of informative wavelengths and quantification of prediction uncertainty. Five alternative methods were all found to perform well with a cross-validated R² of approximately 0.8 and an RMSE of one percentage point, demonstrating feasibility of the proposed imaging setup and computational pipeline.

Exploring uplift modeling with high class imbalance

Uplift modeling refers to individual level causal inference. Existing research on the topic ignores one prevalent and important aspect: high class imbalance. For instance in online environments uplift modeling is used to optimally target ads and discounts, but very few users ever end up clicking an ad or buying. One common approach to deal with imbalance in classification is by undersampling the dataset. In this work, we show how undersampling can be extended to uplift modeling. We propose four undersampling methods for uplift modeling. We compare the proposed methods empirically and show when some methods have a tendency to break down. One key observation is that accounting for the imbalance is particularly important for uplift random forests, which explains the poor performance of the model in earlier works. Undersampling is also crucial for class-variable transformation based models.

Visible Light Spectrum Extraction from Diffraction Images by Deconvolution and the Cepstrum

Accurate color determination in variable lighting conditions is difficult and requires special devices. We considered the task of extracting the visible light spectrum using ordinary camera sensors, to facilitate low-cost color measurements using consumer equipment. The approach uses a diffractive element attached to a standard camera and a computational algorithm for forming the light spectrum from the resulting diffraction images. We present two machine learning algorithms for this task, based on alternative processing pipelines using deconvolution and cepstrum operations, respectively. The proposed methods were trained and evaluated on diffraction images collected using three cameras and three illuminants to demonstrate the generality of the approach, measuring the quality by comparing the recovered spectra against ground truth measurements collected using a hyperspectral camera. We show that the proposed methods are able to reconstruct the spectrum, and, consequently, the color, with fairly good accuracy in all conditions, but the exact accuracy depends on the specific camera and lighting conditions. The testing procedure followed in our experiments suggests a high degree of confidence in the generalizability of our results; the method works well even for a new illuminant not seen in the development phase.

Practical Camera Sensor Spectral Response and Uncertainty Estimation

Knowledge of the spectral response of a camera is important in many applications such as illumination estimation, spectrum estimation in multi-spectral camera systems, and color consistency correction for computer vision. We present a practical method for estimating the camera sensor spectral response and uncertainty, consisting of an imaging method and an algorithm. We use only 15 images (four diffraction images and 11 images of color patches of known spectra to obtain high-resolution spectral response estimates) and obtain uncertainty estimates by training an ensemble of response estimation models. The algorithm does not assume any strict priors that would limit the possible spectral response estimates and is thus applicable to any camera sensor, at least in the visible range. The estimates have low errors for estimating color channel values from known spectra, and are consistent with previously reported spectral response estimates.

Group Factor Analysis

Factor analysis (FA) provides linear factors that describe the relationships between individual variables of a data set. We extend this classical formulation into linear factors that describe the relationships between groups of variables, where each group represents either a set of related variables or a data set. The model also naturally extends canonical correlation analysis to more than two sets, in a way that is more flexible than previous extensions. Our solution is formulated as a variational inference of a latent variable model with structural sparsity, and it consists of two hierarchical levels: 1) the higher level models the relationships between the groups and 2) the lower models the observed variables given the higher level. We show that the resulting solution solves the group factor analysis (GFA) problem accurately, outperforming alternative FA-based solutions as well as more straightforward implementations of GFA. The method is demonstrated on two life science data sets, one on brain activation and the other on systems biology, illustrating its applicability to the analysis of different types of high-dimensional data sources.

Towards brain-activity-controlled information retrieval: Decoding image relevance from MEG signals

We hypothesize that brain activity can be used to control future information retrieval systems. To this end, we conducted a feasibility study on predicting the relevance of visual objects from brain activity. We analyze both magnetoencephalographic (MEG) and gaze signals from nine subjects who were viewing image collages, a subset of which was relevant to a predetermined task. We report three findings: i) the relevance of an image a subject looks at can be decoded from MEG signals with performance significantly better than chance, ii) fusion of gaze-based and MEG-based classifiers significantly improves the prediction performance compared to using either signal alone, and iii) non-linear classification of the MEG signals using Gaussian process classifiers outperforms linear classification. These findings break new ground for building brain-activity-based interactive image retrieval systems, as well as for systems utilizing feedback both from brain activity and eye movements.

Identifying fragments of natural speech from the listener's MEG signals

It is a challenge for current signal analysis approaches to identify the electrophysiological brain signatures of continuous natural speech that the subject is listening to. To relate magnetoencephalographic (MEG) brain responses to the physical properties of such speech stimuli, we applied canonical correlation analysis (CCA) and a Bayesian mixture of CCA analyzers to extract MEG features related to the speech envelope. Seven healthy adults listened to news for an hour while their brain signals were recorded with whole-scalp MEG. We found shared signal time series (canonical variates) between the MEG signals and speech envelopes at 0.5-12 Hz. By splitting the test signals into equal-length fragments from 2 to 65 s (corresponding to 703 down to 21 pieces per the total speech stimulus) we obtained better than chance-level identification for speech fragments longer than 2-3 s, not used in the model training. The applied analysis approach thus allowed identification of segments of natural speech by means of partial reconstruction of the continuous speech envelope (i.e., the intensity variations of the speech sounds) from MEG responses, provided means to empirically assess the time scales obtainable in speech decoding with the canonical variates, and it demonstrated accurate identification of the heard speech fragments from the MEG data.

High Expression of Complement Component 5 (C5) at Tumor Site Associates with Superior Survival in Ewing's Sarcoma Family of Tumour Patients

Background. Unlike in most adult-onset cancers, an association between typical paediatric neoplasms and inflammatory triggers is rare. We studied whether immune system-related genes are activated and have prognostic significance in Ewing's sarcoma family of tumors (ESFTs). Method. Data analysis was performed on gene expression profiles of 44 ESFT patients, 11 ESFT cell lines, and 18 normal skeletal muscle samples. Differential expression of 238 inflammation and 299 macrophage-related genes was analysed by t-test, and survival analysis was performed according to gene expression. Results. Inflammatory genes are activated in ESFT patient samples, as 38 of 238 (16%) inflammatory genes were upregulated (P < 0.001) when compared to cell lines. This inflammatory gene activation was characterized by significant enrichment of macrophage-related gene expression with 58 of 299 (19%) of genes upregulated (P < 0.001). High expression of complement component 5 (C5) correlated with better event-free (P = 0.01) and overall survival (P = 0.004) in a dose-dependent manner. C5 and its receptor C5aR1 expression was verified at protein level by immunohistochemistry on an independent ESFT tumour tissue microarray. Conclusion. Immune system-related gene activation is observed in ESFT patient samples, and prognostically significant inflammatory genes (C5, JAK1, and IL8) for ESFT were identified.

Combined use of expression and CGH arrays pinpoints novel candidate genes in Ewing sarcoma family of tumors

Background

Ewing sarcoma family of tumors (ESFT), characterized by t(11;22)(q24;q12), is one of the most common tumors of bone in children and young adults. In addition to EWS/FLI1 gene fusion, copy number changes are known to be significant for the underlying neoplastic development of ESFT and for patient outcome. Our genome-wide high-resolution analysis aspired to pinpoint genomic regions of highest interest and possible target genes in these areas.

Methods

Array comparative genomic hybridization (CGH) and expression arrays were used to screen for copy number alterations and expression changes in ESFT patient samples. A total of 31 ESFT samples were analyzed by aCGH and in 16 patients DNA and RNA level data, created by expression arrays, was integrated. Time of the follow-up of these patients was 5–192 months. Clinical outcome was statistically evaluated by Kaplan-Meier/Logrank methods and RT-PCR was applied on 42 patient samples to study the gene of the highest interest.

Results

Copy number changes were detected in 87% of the cases. The most recurrent copy number changes were gains at 1q, 2, 8, and 12, and losses at 9p and 16q. Cumulative event free survival (ESFT) and overall survival (OS) were significantly better (P < 0.05) for primary tumors with three or less copy number changes than for tumors with higher number of copy number aberrations. In three samples copy number imbalances were detected in chromosomes 11 and 22 affecting the FLI1 and EWSR1 loci, suggesting that an unbalanced t(11;22) and subsequent duplication of the derivative chromosome harboring fusion gene is a common event in ESFT. Further, amplifications on chromosomes 20 and 22 seen in one patient sample suggest a novel translocation type between EWSR1 and an unidentified fusion partner at 20q. In total 20 novel ESFT associated putative oncogenes and tumor suppressor genes were found in the integration analysis of array CGH and expression data. Quantitative RT-PCR to study the expression levels of the most interesting gene, HDGF, confirmed that its expression was higher than in control samples. However, no association between HDGF expression and patient survival was observed.

Conclusion

We conclude that array CGH and integration analysis proved to be effective methods to identify chromosome regions and novel target genes involved in the tumorigenesis of ESFT.

Simple integrative preprocessing preserves what is shared in data sources

Background

Bioinformatics data analysis toolbox needs general-purpose, fast and easily interpretable preprocessing tools that perform data integration during exploratory data analysis. Our focus is on vector-valued data sources, each consisting of measurements of the same entity but on different variables, and on tasks where source-specific variation is considered noisy or not interesting. Principal components analysis of all sources combined together is an obvious choice if it is not important to distinguish between data source-specific and shared variation. Canonical Correlation Analysis (CCA) focuses on mutual dependencies and discards source-specific "noise" but it produces a separate set of components for each source.

Results

It turns out that components given by CCA can be combined easily to produce a linear and hence fast and easily interpretable feature extraction method. The method fuses together several sources, such that the properties they share are preserved. Source-specific variation is discarded as uninteresting. We give the details and implement them in a software tool. The method is demonstrated on gene expression measurements in three case studies: classification of cell cycle regulated genes in yeast, identification of differentially expressed genes in leukemia, and defining stress response in yeast. The software package is available at .

Conclusion

We introduced a method for the task of data fusion for exploratory data analysis, when statistical dependencies between the sources and not within a source are interesting. The method uses canonical correlation analysis in a new way for dimensionality reduction, and inherits its good properties of being simple, fast, and easily interpretable as a linear projection.

Improved learning of Riemannian metrics for exploratory analysis

We have earlier introduced a principle for learning metrics, which shows how metric-based methods can be made to focus on discriminative properties of data. The main applications are in supervising unsupervised learning to model interesting variation in data, instead of modeling all variation as plain unsupervised learning does. The metrics are derived by approximations to an information-geometric formulation. In this paper, we review the theory, introduce better approximations to the distances, and show how to apply them in two different kinds of unsupervised methods: prototype-based and pairwise distance-based. The two examples are self-organizing maps and multidimensional scaling (Sammon's mapping).