Trjtdtk

Can this transistor (2N2222) take 6 V on emitter-base? Am I reading the datasheet incorrectly?

Are British MPs missing the point, with these 'Indicative Votes'?

Is it a bad idea to plug the other end of ESD strap to wall ground?

Free fall ellipse or parabola?

Car headlights in a world without electricity

What steps are necessary to read a Modern SSD in Medieval Europe?

Finitely generated matrix groups whose eigenvalues are all algebraic

Direct Implications Between USA and UK in Event of No-Deal Brexit

Is a distribution that is normal, but highly skewed, considered Gaussian?

Is it OK to decorate a log book cover?

Arrows in tikz Markov chain diagram overlap

Would a grinding machine be a simple and workable propulsion system for an interplanetary spacecraft?

Is there a rule of thumb for determining the amount one should accept for a settlement offer?

Early programmable calculators with RS-232

Read/write a pipe-delimited file line by line with some simple text manipulation

Incomplete cube

My ex-girlfriend uses my Apple ID to login to her iPad, do I have to give her my Apple ID password to reset it?

Is a linearly independent set whose span is dense a Schauder basis?

Can you teleport closer to a creature you are Frightened of?

logical reads on global temp table, but not on session-level temp table

Why was Sir Cadogan fired?

Man transported from Alternate World into ours by a Neutrino Detector

What is the difference between 'contrib' and 'non-free' packages repositories?

How should I connect my cat5 cable to connectors having an orange-green line?

Activity Recognition of non periodic activites

0

$begingroup$

The data consits of single time series instances of a complex activity that can be divided into single non-periodic activities. The data was collected using mobile phones, holding information about acceleration and motion rotation rate on 3 axis, as well as speed & other sensor data.
I want to detect the single atomic activties. I've already labeled the data completely. However, there are some single activities that are non periodic, while some are periodic (like sitdown/close a door vs. walking). As the time variance of these activities are also very different I dont know which features/approach is suitable for this problem. I expect the classic windowing approach not to work well on here because the windows could miss/split the non-periodic activites.

Do you have any ideas / discussions on this problem? I thought about a 'template matching' approach.

classification time-series activity-recognition

share|improve this question

asked Mar 25 at 9:51

dataddicteddataddicted

112

$endgroup$

add a comment | 

0

$begingroup$

The data consits of single time series instances of a complex activity that can be divided into single non-periodic activities. The data was collected using mobile phones, holding information about acceleration and motion rotation rate on 3 axis, as well as speed & other sensor data.
I want to detect the single atomic activties. I've already labeled the data completely. However, there are some single activities that are non periodic, while some are periodic (like sitdown/close a door vs. walking). As the time variance of these activities are also very different I dont know which features/approach is suitable for this problem. I expect the classic windowing approach not to work well on here because the windows could miss/split the non-periodic activites.

Do you have any ideas / discussions on this problem? I thought about a 'template matching' approach.

classification time-series activity-recognition

share|improve this question

asked Mar 25 at 9:51

dataddicteddataddicted

112

$endgroup$

add a comment | 

$begingroup$

The data consits of single time series instances of a complex activity that can be divided into single non-periodic activities. The data was collected using mobile phones, holding information about acceleration and motion rotation rate on 3 axis, as well as speed & other sensor data.
I want to detect the single atomic activties. I've already labeled the data completely. However, there are some single activities that are non periodic, while some are periodic (like sitdown/close a door vs. walking). As the time variance of these activities are also very different I dont know which features/approach is suitable for this problem. I expect the classic windowing approach not to work well on here because the windows could miss/split the non-periodic activites.

Do you have any ideas / discussions on this problem? I thought about a 'template matching' approach.

classification time-series activity-recognition

share|improve this question

asked Mar 25 at 9:51

dataddicteddataddicted

112

$endgroup$

share|improve this question

asked Mar 25 at 9:51

dataddicteddataddicted

112

share|improve this question

asked Mar 25 at 9:51

dataddicteddataddicted

112

asked Mar 25 at 9:51

dataddicteddataddicted

112

asked Mar 25 at 9:51

dataddicteddataddicted

112