^{1}

^{*}

^{1}

^{1}

^{1}

^{2}

This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (

It is known that parameter selection for data sampling frequency and segmentation techniques (including different methods and window sizes) has an impact on the classification accuracy. For Ambient Assisted Living (AAL), no clear information to select these parameters exists, hence a wide variety and inconsistency across today's literature is observed. This paper presents the empirical investigation of different data sampling rates, segmentation techniques and segmentation window sizes and their effect on the accuracy of Activity of Daily Living (ADL) event classification and computational load for two different accelerometer sensor datasets. The study is conducted using an ANalysis Of VAriance (ANOVA) based on 32 different window sizes, three different segmentation algorithm (with and without overlap, totaling in six different parameters) and six sampling frequencies for nine common classification algorithms. The classification accuracy is based on a feature vector consisting of Root Mean Square (RMS), Mean, Signal Magnitude Area (SMA), Signal Vector Magnitude (here SMV), Energy, Entropy, FFTPeak, Standard Deviation (STD). The results are presented alongside recommendations for the parameter selection on the basis of the best performing parameter combinations that are identified by means of the corresponding Pareto curve.

Ambient Assisted Living (AAL) is currently on the research agenda of many stakeholders worldwide, especially in Western countries, driven mainly by the needs of an aging population and in an attempt to address the demands of care and intervention for the elderly and those who require care. The main areas of interest in Assisted Living (AL) include fall prevention, promotion of independence, as well as ambulation and Activity of Daily Living (ADL) monitoring (for fall detection, activity recognition and classification). The timeliness and accuracy of the classification of ADL activities could have severe consequences if inadequate, especially in the case of an emergency event such as a fall and are therefore essential to provide the elderly with a sense of security and confidence [

Falls and ADL events are generally classified based on the features extracted from segments of the monitoring sensor data and have therefore a significant role in the accuracy of event classification [

The literature review showed that there is no consensus in the selection of parameter combinations which once chosen, are seldom varied by researchers to improve classification results. Therefore, the work described in this paper empirically investigates the influence of sampling frequency (SF), segmentation method (SM), and windows size (WS) on the classification accuracy (CA) and computational load (CL) using two independent datasets (from Bao

Section 2 will highlight existing literature to outline the inconsistency and insufficient justification for parameter selection in ADL classification. This section also presents the process of data acquisition and introduces different segmentation techniques. Section 3 describes the investigation procedure. Section 4 presents the experimental results with a recommendation for parameter combinations, and Sections 5 and 6 present the discussion of results and conclusion.

The acquisition of data is one of the most critical steps in event classification as re-running experiments with test subjects is not always possible. Undersampling leads to loss of information and oversampling can result in information buried in unwanted noise. In the latter case, longer computational time is needed for analysis as more data needs to be processed. The minimum sampling rate _{sampling}_{max}

The highest sampling rate for AAL that the authors found during their research is 512 Hz by [_{samplin}_{g} to be 64 Hz. The authors acknowledge the high frequency sampling rate used by [

One of the challenges of data pre-processing following acquisition consists in deciding which points to actually use in the live stream of data. Several different segmentation methods exist to divide a larger data stream into smaller fit for processing chunks. The selection of the right segmentation technique is crucial, as it immediately impacts on the extracted features used for the ADL classification and the resulting classification accuracy. Therefore even the best classifier performance will be weak when the extracted features are non-differentiable [

Researchers who use fixed size window segmentation methods apply inconsistent window sizes. [

The work of [

This work presented here was based on two different datasets from the literature. The first dataset contains two-axis accelerometer data collected by [

Section 2 showed that sampling rates vary greatly throughout the literature; it also indicated a high use of sampling rates around 50 Hz even though the work of Maurer

The work presented here focuses on three online segmentation techniques: FNSW, FOSW (with four different overlap percentages), and SWAB that were introduced in Section 2.2. As described above, the advantages of these algorithms are that they are online capable, therefore can be used while the data collection is in progress and are simple and intuitive so that they are easily understood.

FNSW is a simple segmentation technique without any data overlap (see _{i}_{Sampling}_{Size}_{Sampling}_{Size}

The FOSW segmentation technique is based on FNSW but includes data overlap (see _{j}

SWAB is the third segmentation technique used as part of the study presented here and was designed by [

As highlighted earlier in Section 2.2, there is no clear recommendation in the published literature on the selection of the window size used for the data segmentation. The authors therefore tested a range of 32 different sizes in the range of 0.5 to 24 s. In the area of 0.5 to 8 s, the size is increased in 0.5 s steps, while thereafter the step size is increased to 1 s. The 0.5 s step size was increased after 8 s because the ADLs under investigation have only a short time frame and computational load was reduced for the experiment. Even though literature showed the use of longer window sizes the aim is to only include single ADLs in each window to achieve the best classification results. The authors' initial research [

The following eight metrics are quite common in the area of ADL classification and therefore used to retrieve the different features of the accelerometer sensor data in this research: Root Mean Square (RMS), Mean, Signal Magnitude Area (SMA), Signal Vector Magnitude (here SMV), Energy, Entropy, FFTPeak, Standard Deviation (STD). These metrics and their significance are discussed below as each individual metric has its own influence in the research field.

RMS has been used to distinguish walking patterns [

The Mean metric (see

The next metric, SMA is used to distinguish between periods of activity and rest in order to identify when the subject is mobilizing and undertaking activities, and when they are immobile [

SMV, normally referred to as Signal Vector Magnitude (SVM) but changed to SMV to avoid confusion with the SVM classifier used, indicates the degree of movement intensity and is an essential metric in fall detection [

Two additional metrics used in this research are Energy and Entropy, which discriminate between types of ADL such as walking, standing still, running, sitting and relaxing [

Another feature that was extracted from the accelerometer data stream is the FFTPeak for each axis. The metric has been used for activity recognition [

The last metric used is Standard Deviation (STD), which has been extensively used for activity recognition [

The software tool Weka implements the algorithms of several different classifiers from which nine were selected based on literature to verify the effects of changes in the described parameters above. [

This section will be split in three subsections to focus on different aspects of the parameter selection problem introduced by the variation in classification methods (CM), sampling frequency (SF), segmentation methods (SM) and window size (WS). The first two analyses use an ANalysis Of VAriance (ANOVA) to investigate the impact of the identified parameters on the two dependent variables classification accuracy (CA, as described in Section 4.1) and computational load (CL, as described in Section 4.2). The analyses have been conducted in SPSS [

This section reports on the impact of the variations of four input parameters (CM, SF, SM, and WS) on classification accuracy (output). During the initial analysis of the ANOVA output, the tested accuracy for the KStar algorithm showed a strong sensitivity to changes in the input parameters (SF, SM, and WS). Its influence on the analysis was significant and superimposed on the results. As a consequence the impact of certain input parameters appeared to be of significance for the classification accuracy, while overall the impact resulted from the sensitivity of the KStar classifier. Therefore the authors decided to exclude KStar in the analysis in order to avoid a misinterpretation of the overall impact of parameters on accuracy.

The ANOVA results (presented in

The table shows the Sums of Squares (a measure for the average variability in the data), Degree of Freedom (df—scores that are free to vary once the mean of the set of scores is known), Mean Square (which is used to estimated the variance), F (F-Ratio represents the indicator for the significance on performance caused by the independent variables instead of chance), and Sig. (indicating the significance level at which the main/two-way interaction effects are significant <0.05 or non-significant >0.05) for all main and two-way interaction effects. The two-way interaction effects outline a changing main effect of one factor for different levels of a second factor and are therefore of higher interest than the main effect alone, if they are identified as being significant. The Sig. column in

The next effect investigated is the interaction between WS and SM in

The last effect under investigation is SF and CM. The graph is not presented, as there is no interaction effect between the different classifiers. The only effect that exists is a minor improvement of accuracy for a change of sampling frequency from 10 to 20 Hz with a nearly constant accuracy thereafter for all classifiers. This correlates with the statement in [

The ANOVA results (presented in

The next effect investigated is the interaction between WS and CM in

The selection of different SF, SM, WS and CM does not only have an impact on the classification accuracy but also on the CL of the system. The CL for the classification of ADL events is based on two main factors. The first one is the data pre-processing and feature extraction step (indicated as _{Time}_{1} in _{Time}_{2} in

_{Time}_{1} depends on the selected SF, SM and WS, excluding any other pre-processing steps such as filtering which is not of interest in this study, while _{Time}_{2} purely depends on the selected CM. For real-time applications, the combination of SM and WS introduces a limitation for certain parameter combinations, leading to the requirement that

The authors therefore conducted an analysis with the CL as the dependent variable to investigate the influence of the four input parameter SF, SM, WS and CM. In the preliminary analysis, one of the levels of the SM input showed to have a high influence on the dependent variable. As before with KStar superimposing on parameters for the accuracy, the SWAB segmentation method increases noticeably the CL as compared to the other methods. Hence, effects that were non-significant before are significant once SWAB is removed as a SM level. Therefore, the analysis will outline the overall input variables without SWAB segmentation method.

The result of the ANOVA is represented in

The interaction effect for segmentation method and classifier method in

The last interaction effect under investigation is WS and SM. The graph in

The result of the ANOVA is represented in

The Source column is sorted based on the Sum of Squares to allow for an easier recognition of the importance of an input parameter. The data highlights that the most important factor for the CL is CM. This is followed by WS, SF and as the least significant parameter SM. For the two-way interaction effect the four significant combinations are WS and CM, SM and CM, SF and WS, followed by SM and WS. The interaction effect SF and SM that was significant earlier, is non-significant for this dataset.

The interaction effect for segmentation method and classifier in

The last interaction effect under investigation is windows size and segmentation method. The graph in

Based on the parameter influence described in Sections 4.1 and 4.2, the inevitable question still stands: what is the best parameter selection for a given requirement? The answer, however, depends strongly on the preference with respect to classification performance, e.g., is the best accuracy required or are there limitations to CL. Therefore, a set of well-performing parameter sets based on the trade-off between accuracy and CL were identified. For the given dataset certain parameter combinations will achieve a similar CA but will require different CL and vice versa. When plotted in a graph, such as presented in

Short WS have a significant influence in this dataset, as both datasets include parameter combinations with 1.5 s or less. The dominating classifiers are Naïve Bayes, and KNN. This is a slight variation compared to the 10 Hz limited Bao

One of the main problems in AAL is the availability (or the lack thereof) of test subjects, as compared to clinical trials, where subjects can reach into the thousands. In [

In summary the outputs of the work presented here, are listed below:

The importance of parameters for CA ranked in order of decreasing influence is CM, SM, WS and SF;

The impact of WS is different for both datasets;

Increased segmentation overlap improves CA;

The influence of SWAB on CA is different in both datasets;

SF above 10 Hz has only a minor improvement on CA;

CL behaves the same for both dataset;

The importance of parameters for CL ranked in order of decreasing influence is CM, WS, SF and SM;

Some dominant parameter combinations of the Pareto curve are similar for both datasets;

Higher CL does not automatically result in higher CA.

The following discussion will look into the results of the ANalysis Of VAriance (ANOVA) for CA and CL and finish with the dominant parameter points of the Pareto curves. The two-way interaction effect between SM and CM highlights for both datasets that FOSW with 90% overlap results in the best CA. From FNSW (no overlap) to FOSW with 90% overlap, both datasets show that more overlap improves the CA. A possible reason for this is that the increase in overlap allows for a bigger training set and has the lowest loss of information, in the range of investigated SM. The results for SWAB are mixed. For the Bao

For CL, both datasets show the same behavior for the two-way interaction effects. For the three interaction effects including WS (WS and CM, WS and SM, WS and SF) similar behavior is observable. A shorter WS results in a lower CL, while a longer WS will increase the CL. This effect is lowest for WS and CM and highest for WS and SM. The interaction effect between SM and CM highlights no significant change for any classifier besides SMO. SMO is the only classifier that can reduce the CL with an increased segmentation overlap.

The authors used ANOVA to quantify the influence of the different parameters on the CA and CL. They have also used a Pareto curve based approach to highlight dominant parameter combinations for “optimum” achievable performance (optimality being decided by the user in a given context/application).

This paper has presented a new instrument to help select data capture and processing parameters for the recognition of Activities of Daily Living (ADL). A review of the literature uncovered a lack of consensus in terms of the selection of sampling frequency, segmentation method and window size, and classifier method for the recognition of ADL. The impact of the sampling frequency (six levels), segmentation method (three segmentation algorithms with different parameters resulting in six different levels) and segmentation window size (32 levels) on the classification accuracy and computational load of a set of commonly used classifiers (nine levels) has been investigated. This has involved experimenting with two datasets, containing 20 and three test subjects, respectively, and analysis of the resulting data using ANalysis Of VAriance (ANOVA). The analysis showed that the choice of classifier method is the most important parameter followed by the segmentation method, window size and finally sampling frequency. It also showed that in the case of computational load the parameters ranked in order of decreasing influence are classifier method, window size, sampling frequency, and segmentation method. The results have been presented graphically using a Pareto curve, which highlighted two dominant classifiers for both datasets (KNN, Naïve Bayes). The Pareto curve did not show matching dominant points in both datasets, however, it showed that combinations of three out of the four factors (CM, SM, SF) are likely to result in dominant points. The authors have suggested that the Pareto curve is a good instrument which can be used to select sets of parameters based on their impact on classification accuracy and computational load and resolve trade-off issues.

As part of their future work in the general area of AAL, the authors plan to investigate a number of issues specific to the findings presented in this paper. An important point of interest is the identification of the reasons behind the inconsistency between the two datasets used in terms of the influence of WS on the classification accuracy. A possible influential factor, not considered in the present work, is the nature of the ADL itself. It might be necessary to adjust the WS parameter with regards to the expected ADLs in the dataset; [

Sebastian D. Bersch, Ifeyinwa E. Achumba and Djamel Azzi designed the experiment, Sebastian D. Bersch performed the experiments, Sebastian D. Bersch, Jana Ries, Djamel Azzi and Rinat Khusainov analyzed the data, Sebastian D. Bersch, Jana Ries, Ifeyinwa E. Achumba, Djamel Azzi and Rinat Khusainov prepared the manuscript.

The authors declare no conflict of interest.

Pre-steps before ADL.

Explanation of segmentation method. (

Explanation of segmentation method SWAB.

Parameter combinations for each classifier.

Two-way interaction effect for SM and CM.

Two-way interaction effect for WS and CM.

Two-way interaction effect for WS and SM.

Two-way interaction effect for WS and SF.

Two-way interaction effect for WS and SM.

Two-way interaction effect for SM and CM.

Two-way interaction effect for WS and CM.

Timing factor for computational load.

Two-way interaction effect for WS and CM.

Two-way interaction effect for SM and CM.

Two-way interaction effect for WS and SF.

Two-way interaction effect for WS and SM.

Two-way interaction effect for WS and CM.

Two-way interaction effect for SM and CM.

Two-way interaction effect for WS and SF.

Two-way interaction effect for WS and SM.

Explanation of the Pareto curve.

Dominant points on the Pareto curve.

Dominant points on the Pareto curve.

Dominant points on the Pareto curve for both datasets.

Inconsistency in sampling rates and segmentation windows for AAL.

Huynh [ |
512 | 0.25, 0.5, 1, 2, 4 | FNSW, FOSW 50%, FOSW 75%, FOSW 80.5%, FOSW 93.75% | Walking, Standing, Jogging, Skipping, Hopping, Riding Bus | |

Sekine [ |
256 | Subjects 11; Age 69.3 ± 5.6 years; Height 1.54 ± 0.078 m; Weight 50.4 ± 9.6 kg | Walking | ||

Bao [ |
76.25 | 6.7 | FOSW 50% | Subjects: 13 male, 7 female; Age 17–48 years | Walking, Sitting & Relaxing, Standing Stil, Watching TV, Running, Stretching, Scrubbing, Folding Laundry, Brushing teeth, Riding Elevator, Walking Carrying items, Working on Computer, Eating or Driniking, Reading, Bicycling, Strength Training, Vacuuming, Lying Down & Relaxing, Climbing Stairs, Riding Escalator |

Preece [ |
64 | 2 and 3 | FOSW 50% | Subjects: 10 male, 10 female; Age 31 ± 7 years; Height 1.71 ± 0.07 m; Weight 68 ± 10 kg; BMI 24 ± 3 | Walking, Walking Upstairs, Walking Downstairs, Hopping on Left Leg, Hopping on Right Leg, Jumping |

Wang [ |
50 | 2.56 | FOSW 50% | Subjects: 39 male, 12 female; Age 21–64 years; Height 1.53–188 m; Weight 42–94 kg | Walking, Walking Slope Up, Walking Slope Down, Walking Stairs Up, Walking Stairs Down |

Casale [ |
52 | 1 | FOSW 50% | Subjects: 11 male, 3 female | Walking Stairs Up, Walking Stairs Down, Walking, Talking, Staying Standing, Working at Computer |

Ravi [ |
50 | 5.12 | FOSW 50% | Subjects 2 | Standing, Walking, Running, Walking Stairs Up, Walking Stairs Down, Situps, Vacuuming, Brushing Teeth |

Pärkkä [ |
50 | 5 | Subjects 7; median (range); 27 years (4–37); Height 180 (92–187) | Lying, Sitting, Standing, Walking, Bicycling, Running | |

Maurer [ |
50 | 4 | FOSW 92% | Subjects 6 | Sitting, Standing, Walking, Walking Stairs Up, Walking Stairs Down, Running |

Antonsson [ |
1–30 | Subjects 12 | Walking (Gait) | ||

Bouten [ |
20 | Subjects: 13 male; Age 27 ± 4 years; Height 1.83 ± 0.07 m; Weight 77 ± 12 kg | Sedentary Activities, Household Activities, Walking | ||

Gjoreski [ |
5 | 1.4 | Standing, Lying, Sitting, On all fours, Sitting on the Ground, Going Down, Standing Up | ||

Nyan [ |
256 | 2 | Subjects 22; Age 20–45 years; Height 1.67–1.94 m; Weight 45–93 kg | Walking, Walking Upstairs, Walking Downstairs | |

Kasteren [ |
60 | FNSW | Subject 1 | Leaving House, Toileting, Showering, Sleeping, Breakfast, Dinner, Drink | |

Patterson [ |
74 | Subject 1 | Using Bathroom, Making Oatmeal, Making Soft-Boiled Eggs, Preparing Orange-Juice, Making Coffee, Making Tea, Making or Answering a Phone Call, Taking out the Trash, Setting the Table, Eating Breakfast, Clearing Table | ||

Pietka [ |
FNSW, FOSW, SAX, SM | ||||

Keogh [ |
FNSW, FOSW, Bup, SWAB | ||||

Chu [ |
RbW | ||||

Kozina [ |
Dwin | ||||

Ortiz Laguna [ |
VSW |

ANOVA output for the CA as the dependent variable (Tests of Between-Subjects Effects. Dependent Variable: CA).

Corrected Model | 3,257,844 ^{a} |
670 | 4,862 | 260 | 0.000 |

Intercept | 1,503,713,645 | 1 | 1,503,713,645 | 80,275,936 | 0.000 |

SF | 116,380 | 5 | 23,276 | 1,243 | 0.000 |

WS | 216,554 | 31 | 6,986 | 373 | 0.000 |

SM | 650,201 | 5 | 130,040 | 6,942 | 0.000 |

CM | 1,961,904 | 7 | 280,272 | 14,962 | 0.000 |

SF * SM | 5,924 | 25 | 237 | 13 | 0.000 |

SF * CM | 7,591 | 35 | 217 | 12 | 0.000 |

SF * WS | 36,024 | 155 | 232 | 12 | 0.000 |

SM * WS | 60,930 | 155 | 393 | 21 | 0.000 |

WS * CM | 92,164 | 217 | 425 | 23 | 0.000 |

SM * CM | 110,091 | 35 | 3,145 | 168 | 0.000 |

Error | 3,439,779 | 183,633 | 19 | ||

Total | 1,510,410,864 | 184,304 | |||

Corrected Total | 6,697,622 | 184,303 |

R Squared = 0.486 (Adjusted R Squared = 0.485).

ANOVA output for the CA as the dependent variable (Tests of Between-Subjects Effects. Dependent Variable: CA).

Corrected Model | 1,785,658 ^{a} |
670 | 2,665 | 45 | 0.000 |

Intercept | 147,805,613 | 1 | 147,805,613 | 2,587,861 | 0.000 |

SF | 9,501 | 5 | 1,900 | 33 | 0.000 |

WS | 183,628 | 31 | 5,923 | 104 | 0.000 |

SM | 639,572 | 5 | 127,914 | 2,240 | 0.000 |

CM | 759,536 | 7 | 108,505 | 1,900 | 0.000 |

SF * SM | 778 | 25 | 31 | 0.545 | 0.968 |

SF * CM | 1,445 | 35 | 41 | 0.723 | 0.886 |

SF * WS | 4,898 | 155 | 32 | 0.553 | 1.000 |

WS * CM | 25,038 | 217 | 115 | 2 | 0.000 |

SM * CM | 70,397 | 35 | 2,011 | 35 | 0.000 |

SM * WS | 90,865 | 155 | 586 | 10 | 0.000 |

Error | 1,540,791 | 26,977 | 57 | ||

Total | 151,132,062 | 27,648 | |||

Corrected Total | 3,326,449 | 27,647 |

R Squared = 0.537 (Adjusted R Squared = 0.525).

ANOVA output for the CL as the dependent variable (Tests of Between-Subjects Effects. Dependent Variable: CL).

Corrected Model | 1.316 ^{a} |
626 | 0.002 | 222 | 0.000 |

Intercept | 7.755 | 1 | 7.755 | 818,571 | 0.000 |

SM | 0.013 | 4 | 0.003 | 355 | 0.000 |

SF | 0.104 | 5 | 0.021 | 2,202 | 0.000 |

WS | 0.330 | 31 | 0.011 | 1,122 | 0.000 |

CM | 0.515 | 7 | 0.074 | 7,769 | 0.000 |

SF * CM | 0.000 | 35 | 3.587 × 10^{−6} |
0.379 | 1.000 |

SF * SM | 0.001 | 20 | 5.456 × 10^{−5} |
5.759 | 0.000 |

SM * WS | 0.034 | 124 | 0.000 | 29 | 0.000 |

SF * WS | 0.082 | 155 | 0.001 | 56 | 0.000 |

SM * CM | 0.103 | 28 | 0.004 | 390 | 0.000 |

WS * CM | 0.134 | 217 | 0.001 | 65 | 0.000 |

Error | 1.449 | 152,957 | 9.474 × 10^{−6} |
||

Total | 10.521 | 153,584 | |||

Corrected Total | 2.766 | 153,583 |

R Squared = 0.476 (Adjusted R Squared = 0.474).

ANOVA output for the CL as the dependent variable (Tests of Between-Subjects Effects. Dependent Variable: CL).

Corrected Model | 0.358 ^{a} |
626 | 0.001 | 93 | 0.000 |

Intercept | 1.525 | 1 | 1.525 | 247,087 | 0.000 |

SM | 0.008 | 4 | 0.002 | 328 | 0.000 |

SF | 0.021 | 5 | 0.004 | 671 | 0.000 |

WS | 0.090 | 31 | 0.003 | 470 | 0.000 |

CM | 0.112 | 7 | 0.016 | 2,595 | 0.000 |

SF * SM | 0.000 | 20 | 7.884 × 10^{−6} |
1 | 0.181 |

SF * CM | 0.000 | 35 | 7.236 × 10^{−6} |
1 | 0.223 |

SM * WS | 0.006 | 124 | 4.866 × 10^{−5} |
8 | 0.000 |

SF * WS | 0.011 | 155 | 6.966 × 10^{−5} |
11 | 0.000 |

SM * CM | 0.050 | 28 | 0.002 | 290 | 0.000 |

WS * CM | 0.060 | 217 | 0.000 | 45 | 0.000 |

Error | 0.138 | 22,413 | 6.172 × 10^{−6} |
||

Total | 2.021 | 23,040 | |||

Corrected Total | 0.496 | 23,039 |

R Squared = 0.721 (Adjusted R Squared = 0.714).