**3. Materials**

A sensor network capable of acquiring the ECG, TEB and EDA signals has been designed, in order to obtain a database of signals to be used in this study. The complete system acquires all the signals described in the literature, that have been mentioned above, which is explained in detail in [25,60]. To acquire the multimodal biosignals a set of sensorized garments were used, which are shown in Figure 1. A glove to acquire EDA measurement in hand, a bracelet to acquire EDA measurement in the arm and a vest to acquire ECG and TEB. These garments are connected to the measurement devices shown in Figure 2. The glove and the bracelet are connected to the device called GSR, which acquires the EDA signal and the vest is connected to the vest through a recorder called ECGZ2, which acquires ECG and TEB signals. The ECGZ2 is capable of sampling each signal with a different sampling frequency. For the TEB and EDA, the sampling frequency is 100 Hz and for the ECG, the sampling frequency is 250 Hz.

Measurements were collected from *k* = 40 subjects, students and climbers aged 20 to 49, including 12 females and 28 males. The total duration of the complete experiment was approximately 90 min per subject. All of the experiments were performed under the conditions of respect for individual rights and ethical principles that govern biomedical research involving human beings, and written informed consent was obtained from all participants. The experiments were approved by the Research Ethics

Committee at the University of Alcala, and the protocol number assigned to this experiment is CEI: 2013/025/201130624.

**Figure 1.** Recording Devices: (**A**) Electrocardiogram (ECG) and Thoracic Electrical Bioimpedance (TEB) device, (**B**) electrodermal activity (EDA) device. (**A**) Glove to acquire EDA signal in hand; (**B**) Arm bracelet to acquire EDA signal. (**C**) Vest to acquire ECG and TEB signals prior published in [25] under license CC by 4.0.

**Figure 2.** Devices: (**A**) ECGZ2 device, (ECG and TEB recorder prior published in [25] under license CC by 4.0); (**B**) GSR device (EDA recorder published in [60] under license CC by 4.0).

As was stated above, there were four different activities to be recognized: emotional activity, mental activity, physical activity, and neutral activity (resting).

In order to elicit the different activities, we have used a segmen<sup>t</sup> documentary called "Earth" to induce Neutral Activity. In order to elicit emotional activity, we used a set of segments extracted from several validated movies [61]. "American History X" (1998) by Savoy Pictures [62], "I am legend" (2007) by Warner Bross [63], "Life is beautiful" (1997) by Miramax [64,65] and "Cannibal Holocaust" (1980) by F.D. Cinematografica [66]. The mental activity was elicited using a set of games based on mental arithmetic and playing the well-known game "Tetris", used several times to elicit mental activity [67].

The designed activity recognition system had to take a decision every 10 s, and each individual generated 28 time slots of each activity (the database is balanced). Thus, the total number of patterns (decisions) for this analysis was 4480, and each class is composed of 1120 different patterns.

In the present analysis, we have used four different activities:


The database particular characteristics can be found in [25]. The full dataset can also be downloaded from the Supplementary Information included in the paper.
