A Framework for Learning Analytics Using Commodity Wearable Devices

. 2017 Jun 14;17(6):1382. doi: 10.3390/s17061382

Algorithm 1 Student state distinction (SSD) algorithm.

Input:

N_{w}

N_{h}

N_{m}

N_{s}

Φ_{h}

, time period length T, threshold

θ_{n}

and

θ_{h}

, weights

w_{w}

w_{h}

w_{m}

w_{s}

Output: Inferred learner state for current time period.

{\bar{N}}_{T} = \frac{w_{w} N_{w} + w_{h} N_{h} + w_{m} N_{m} + w_{s} N_{s}}{T}

;

σ_{T} = s t d (Φ_{h})

;

3: if

\bar{N} > θ_{n}

and

σ_{T} > θ_{h}

then

4: Infer learner active state during current time period;

5: else if

\bar{N} < θ_{n}

and

σ_{T} < θ_{h}

then

6: Infer learner inactive state during current time period;

7: else

8: Infer unclear state during current time period;