Abstract
Variable selection involves selecting truly important predictors from p-dimensional multivariate functional predictors in functional predictive models. In this paper, a variable selection method is designed for scalar-on-function predictions entangled with nonlinear joint associations among scalar response and multiple functional predictors. First, a nonparametric functional nonlinear conditional correlation coefficient, namely, the FunNCC coefficient, is proposed to measure complex dependencies, including the nonmonotonic marginal dependence, along with the conditional associations of redundancy, complement, and interaction. Then, a model-free feature ordering and selection method is designed, where the FunNCC is utilized to rank relevance, enabling the selection of a subset of predictors with the strongest joint dependence. Since this method allows for quantitatively evaluating the contributions of predictors in explaining responses, it achieves moderate model interpretability. Finally, extensive simulation studies and two real-data cases involving air pollution regression and hand gesture recognition are conducted to evaluate the finite sample performance of the proposed method, and the results show that the proposed FunNCC and variable selection methods outperform state-of-the-art baselines.
Similar content being viewed by others
Notes
The link for the dataset is https://archive.ics.uci.edu/ml/datasets/Beijing+Multi-Site+Air-Quality+Data.
The link to this dataset is http://ninapro.hevs.ch/data1.
References
Amma C, Krings T, Böer J, Schultz T (2015) Advancing muscle–computer interfaces with high-density electromyography. In: Proceedings of the 33rd annual ACM conference on human factors in computing systems, pp 929–938, Seoul Republic of Korea. ACM
Aneiros G, Novo S, Vieu P (2022) Variable selection in functional regression models: a review. J Multivar Anal 188:104871
Atzori M, Gijsberts A, Castellini C, Caputo B, Hager A-GM, Elsig S, Giatsidis G, Bassetto F, Müller H (2014) Electromyography data for non-invasive naturally-controlled robotic hand prostheses. Sci Data 1(1):140053
Azadkia M, Chatterjee S (2021) A simple measure of conditional dependence. Ann Stat 49(6):3070–3102
Bach FR, Jordan MI (2003) Kernel independent component analysis. J Mach Learn Res 3:1–48
Berrendero JR, Cuevas A, Torrecilla JL (2016) The mrmr variable selection method: a comparative study for functional data. J Stat Comput Simul 86(5):891–907
Blanquero R, Carrizosa E, Jiménez-Cordero A, Martín-Barragán B (2019) Variable selection in classification for multivariate functional data. Inf Sci 481:445–462
Brockhaus S, Fuest A, Mayr A, Greven S (2018) Signal regression models for location, scale and shape with an application to stock returns. J R Stat Soc Ser C Appl Stat 67(3):665–686
Chatterjee S (2021) A new coefficient of correlation. J Am Stat Assoc 116(536):2009–2022
Chen H, Zhang Y, Zhou D, Liu H (2020) Improving gesture recognition by bidirectional temporal convolutional networks. In: Qian J, Liu H, Cao J, Zhou D (eds) Robotics and rehabilitation intelligence, communications in computer and information science, pp 413–424. Springer, Singapore
Cheng Y, Shi JQ, Eyre J (2020) Nonlinear mixed-effects scalar-on-function models and variable selection. Stat Comput 30(1):129–140
Collazos JAA, Dias R, Zambom AZ (2016) Consistent variable selection for functional regression models. J Multivar Anal 146:63–71
Dziak JJ, Coffman DL, Reimherr M, Petrovich J, Li R, Shiffman S, Shiyko MP (2019) Scalar-on-function regression for predicting distal outcomes from intensively gathered longitudinal data: interpretability for applied scientists. Stat Surv 13:150
Fan Y, James GM, Radchenko P (2015) Functional additive regression. Ann Stat 43(5):2296–2325. https://doi.org/10.1214/15-AOS1346
Febrero-Bande M, González-Manteiga W, de la Fuente MO (2019) Variable selection in functional additive regression models. Comput Stat 34(2):469–487
Feix T, Romero J, Schmiedmayer H-B, Dollar AM, Kragic D (2016) The grasp taxonomy of human grasp types. IEEE Trans Hum Mach Syst 46(1):66–77
Feng S, Zhang M, Tong T (2022) Variable selection for functional linear models with strong heredity constraint. Ann Inst Stat Math 74(2):321–339
Friedman JH, Rafsky LC (1983) Graph-theoretic measures of multivariate association and prediction. Ann Stat 11(2):377–391. https://doi.org/10.1214/aos/1176346148
Fuchs K, Scheipl F, Greven S (2015) Penalized scalar-on-functions regression with interaction term. Comput Stat Data Anal 81:38–51
Fukumizu K, Bach FR, Gretton A (2007) Statistical consistency of kernel canonical correlation analysis. J Mach Learn Res 8:361–383
Gertheiss J, Goldsmith J, Crainiceanu C, Greven S (2013) Longitudinal scalar-on-functions regression with application to tractography data. Biostatistics 14(3):447–461
Górecki T, Krzyśko M, Wołyński W (2020) Independence test and canonical correlation analysis based on the alignment between kernel matrices for multivariate functional data. Artif Intell Rev 53(1):475–499
Huang L, Zhao J, Wang H, Wang S (2016) Robust shrinkage estimation and selection for functional multiple linear model through lad loss. Comput Stat Data Anal 103:384–400
Jarque-Bou NJ, Scano A, Atzori M, Müller H (2019) Kinematic synergies of hand grasps: a comprehensive study on a large publicly available dataset. J Neuroeng Rehabil 16(1):63
Jung P-G, Lim G, Kim S, Kong K (2015) A wearable gesture recognition device for detecting muscular activities based on air-pressure sensors. IEEE Trans Ind Inform 11(2):485–494. https://doi.org/10.1109/TII.2015.2405413
Kim KK, Kim M, Pyun K, Kim J, Min J, Koh S, Root SE, Kim J, Nguyen B-NT, Nishio Y, Han S, Choi J, Kim C-Y, Tok JB-H, Jo S, Ko SH, Bao Z (2023) A substrate-less nanomesh receptor with meta-learning for rapid hand task recognition. Nat Electron 6(1):64–75. https://doi.org/10.1038/s41928-022-00888-7
Kong D, Xue K, Yao F, Zhang HH (2016) Partially functional linear regression in high dimensions. Biometrika 103(1):147–159
Lai T, Zhang Z, Wang Y (2021) A kernel-based measure for conditional mean dependence. Comput Stat Data Anal 160:107246
Lee CE, Zhang X, Shao X (2020) Testing conditional mean independence for functional data. Biometrika 107(2):331–346. https://doi.org/10.1093/biomet/asz070
Li Y, Qiu Y, Yuhang X (2022) From multivariate to functional data analysis: fundamentals, recent developments, and emerging areas. J Multivar Anal 188:104806
Lian H (2013) Shrinkage estimation and selection for multiple functional regression. Stat Sin 23:51–74
Liu Y, Zeng B, Zhang T, Jiang L, Liu H, Ming D (2021) Quantitative investigation of hand grasp functionality: hand joint motion correlation, independence, and grasping behavior. Appl Bionics Biomech 1–14:2021
Luis TJ, Jose B, Antonio C (2016) Variable selection in functional data classification: a maxima-hunting proposal. Stat Sin 26:619–638. https://doi.org/10.5705/ss.202014.0014
Matsui H (2019) Sparse group lasso for multiclass functional logistic regression models. Commun Stat Simul Comput 48(6):1784–1797
Matsui H, Konishi S (2011) Variable selection for functional regression models via the regularization. Comput Stat Data Anal 55(12):3304–3310
Park SY, Xiao L, Willbur JD, Staicu A-M, Jumbe NL (2018) A joint design for functional data with application to scheduling ultrasound scans. Comput Stat Data Anal 122:101–114
Ramsay JO, Silverman BW (2002) Applied functional data analysis: methods and case studies. Springer series in statistics. Springer, New York
Ramsay JO, Silverman BW (2005) Functional data analysis. Springer series in statistics. Springer, New York
Ramsay J, Hooker G, Graves S (2009) Functional data analysis with R and MATLAB. Springer, New York
Redd A (2012) A comment on the orthogonalization of B-spline basis functions and their derivatives. Stat Comput 22(1):251–257
Sang P, Kashlak AB, Kong L (2022) A reproducing kernel hilbert space framework for functional classification. J Comput Graph Stat 32(3):1000–1008. https://doi.org/10.1080/10618600.2022.2138407
Soham S, Ghosh Anil K (2018) Some multivariate tests of independence based on ranks of nearest neighbors. Technometrics 60(1):101–111
Shao X, Zhang J (2014) Martingale difference correlation and its use in high-dimensional variable screening. J Am Stat Assoc 109(507):1302–1318
Shi H, Drton M, Han F (2021) On Azadkia–Chatterjee’s conditional dependence coefficient. arXiv e-prints
Stein EM, Shakarchi R (2005) Real analysis: measure theory, integration, and Hilbert spaces. Number v. 3 in Princeton lectures in analysis. Princeton University Press, Princeton
Sun J, Liao H, Upadhyaya BR (2014) A robust functional-data-analysis method for data recovery in multichannel sensor systems. IEEE Trans Cybern 44(8):1420–1431
Sun Y, Liu Z, Wang W (2022) Subgroup analysis for the functional linear model. arXiv e-prints
Tai APK, Mickley LJ, Jacob DJ (2010) Correlations between fine particulate matter (pm2.5) and meteorological variables in the united states: implications for the sensitivity of pm25 to climate change. Atmos Environ 44(32):3976–3984
Thind B, Multani K, Cao J (2023) Deep learning with functional inputs. J Comput Graph Stat 32(1):171–180. https://doi.org/10.1080/10618600.2022.2097914
Torrecilla JL, Suárez A (2016) Feature selection in functional data classification with recursive maxima hunting. In: Proceedings of the 30th international conference on neural information processing systems, NIPS’16, pp 4842-4850, Red Hook, NY, USA. Curran Associates Inc
Usset J, Staicu A-M, Maity A (2016) Interaction models for functional regression. Comput Stat Data Anal 94:317–329
Wan Y, Xu M, Huang H, Xi CS (2021) A spatio-temporal model for the analysis and prediction of fine particulate matter concentration in Beijing. Environmetrics 32(1):e2648. https://doi.org/10.1002/env.2648
Wan J, Chen H, Li T, Huang W, Li M, Luo C (2022) R2ci: information theoretic-guided feature selection with multiple correlations. Pattern Recogn 127:108603
Yang K, Manjin X, Yang X, Yang R, Chen Y (2021) A novel emg-based hand gesture recognition framework based on multivariate variational mode decomposition. Sensors 21(21):7002
Yao F, Muller H-G (2010) Functional quadratic regression. Biometrika 97(1):49–64
Yu W, Wade S, Bondell HD, Azizi L (2022) Nonstationary Gaussian process discriminant analysis with variable selection for high-dimensional functional data. J Comput Graph Stat 1–13
Zhu H, Li R, Zhang R, Lian H (2020) Nonlinear functional canonical correlation analysis via distance covariance. J Multivar Anal 180:104662
Acknowledgements
The first author extends sincere gratitude to Professor Bo Liu at the Chinese Academy of Sciences and Dr. Lei Zhu at Beihang University for their generous support throughout the experiments conducted on real data. Additionally, she acknowledges Dr. Xiaoping Liang at the University of Tokyo for insightful discussions on hand gesture recognition. This work was supported by grants from the National Natural Science Foundation of China (Grant Nos. 72021001 and 11701023).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A: FOFCD-Back and FOFCD-Step
Pseudocodes of FOFCD-Back and FOFCD-Step are shown in Algorithms A1 and A2, respectively.
FOFCD-Back starts with the full set of input predictors to be selected. All the elements in the selected set \(j\in {\hat{\mathcal{S}}}\) are subsequently assessed by \(\text{RelS}( j,{\textbf{Y}}\!\mid \!({\hat{\mathcal{S}}}\backslash \{j\}))\); i.e., their conditional dependencies on the response given other elements in the selected set. Predictors with lower-than-\(\delta\) RelS are removed from \({\hat{\mathcal{S}}}\). The process is iterated until the stopping criterion is satisfied. The stopping criterion of FOFCD-Back is that the cardinality of the selected subset \(\left| \hat{S} \right|\) reaches the predetermined lower bound \(d_{\min }\) or \(\text{RelS}( j,{\textbf{Y}}\!\mid \!({\hat{\mathcal{S}}}\backslash \{j\}))\) of all \(j\in {\hat{\mathcal{S}}}\) are no less than the relevant threshold \(\delta\).
FOFCD-Step starts with an empty selected subset and a candidate set that equals the index set of inputs. In each iteration, the most relevant predictor is selected from \({\mathcal{V}}\), and its index is put into \({\hat{\mathcal{S}}}\) to form a new selected subset. Every time \({\hat{\mathcal{S}}}\) is updated, all the elements in the selected set \(j\in {\hat{\mathcal{S}}}\) are assessed, and the conditionally independent predictor with the smallest RelS is returned to the candidate set. The process is iterated until the stopping criterion is satisfied, and the stopping criterion of FOFCD-Step is the same as that of FOFCD-FW.
Appendix B: Supplementary information on Ninapro DB1
1.1 Appendix B.1: Sensor positions of Ninapro DB1
The sensor positions are shown in Fig. 5. sEMG signals reveal muscle activity using 10 surface electromyography electrodes. Eight electrodes were evenly placed around the forearm at a constant distance from the radiohumeral joint below the elbow joint, and the other two electrodes were placed on the flexor major and extensor muscles of the forearm. Kinematic signals were collected using CyberGlove II with 22 sensors, which measure angular changes between pairs of hand joints. Every sample is linked to an accurate timestamp. All the data streams were linearly interpolated to the maximum recorded frequency of 100 Hz to eliminate the differences in the sampling rates of the sEMG and kinematic signals.
1.2 Appendix B.2: Variable selection results
The variable selection results for FOFCD and fLARS are listed in Table 13, including the results for the three subclasses and the whole Ninapro DB1. For FOFCD, a subset of signals from Gloves 1–22 was selected for all datasets. All sEMG signals were retained by FOFCD, except a few evenly placed sEMG signals that were removed on DB1_E2. Gloves 2, 3, 4 and 11, and most sEMG signals were important predictors in all the datasets.
1.3 Appendix B.3: Post hoc interpretation
Feix et al. (2016) classified 20 daily movements into three oppositions, namely palm, pad, and side, according to differences in force directions between the hand and objects, as shown in Fig. 6a, b, respectively. In these figures, virtual fingers (VF) were defined as several fingers that work together as a unit for certain tasks. In Fig. 6a, VF 1 of the palm was the palm, and the third or forth fingers acted as VF 2. For both pad and side, VF 1 were thumb, but VF 2 at each position varied greatly. In the example shown in Fig. 6b, the forefinger opposed the thumb and acted as VF 2.
A visual analysis of the variable selection results on these oppositions was as follows.
-
1.
Movements of the palm mainly involve holding objects by squeezing all powerful fingers perpendicularly toward the palm, e.g., by grabbing a hammer or screwdriver. Predictors selected by FOFCD were mostly located at the tips of middle finger and ring finger (Gloves 10 and 14), as well as the palm (Gloves 4, 5, 12, 15, and 16), and only Glove 3 on thumb was selected, as is shown in Fig. 6d.
-
2.
Pad usually involves the thumb and other fingers, with contact near or at the fingertips and sometimes include contact with palmar surfaces, e.g., by holding a needle or a small ball. The predictors selected by FOFCD were mostly located at the tips of the forefinger and middle finger and at the wrist. All signals in the thumb were also retained, as shown in Fig. 6e.
-
3.
Side mainly involves the tip of the thumb and transverse sides of the other four fingers, e.g., holding a key between the thumb and the radial side of the remaining fingers or holding a cigarette between the fingers. In FOFCD, predictors were mostly located at the five fingers and less often at the palm than the palm opposition, as shown in Fig. 6f.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Wang, K., Wang, H., Wang, S. et al. Variable selection for multivariate functional data via conditional correlation learning. Comput Stat (2024). https://doi.org/10.1007/s00180-024-01489-y
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00180-024-01489-y