Rejected

:speech_balloon:

Disable some Activity types

I frequently have bicycle come up as the picked activity when in fact I am either on public transit or in a car. Given I don’t own a bicycle, it would be helpful to disable certain activities if there is no or very little possibility for them being the correct activity.

3 votes

Tagged as Suggestion

Suggested 14 August 2019 by user Chester Ismay

Moved into Rejected 14 August 2019

  • Sign in to comment and vote. Sign in by email
  • 14 August 2019 Chester Ismay suggested this task

  • 14 August 2019 Matt Greenfield moved this task into Rejected

  • avatar

    Hi!

    Arc already does this automatically. Every time you confirm or correct an activity type in your timeline, Arc includes that information in its Machine Learning models, and learns your personal modes of transport, times of day, common routes, and many other details.

    That learning is used when detecting the probabilities for each activity type for each trip in your timeline. Modes of transport that you never use will have the lowest probability, and will be placed at the bottom of the list. Your common routes will be auto detected as the mode of transport that you typically use for that route, at that time of day (and taking into account a whole bunch of other learned factors as well).

    As long as you confirm or correct the trip items that Arc requests confirmation for, and correct the ones that it has guessed wrong, Arc will do what you want automatically. So there is no need to disable any activity types.

    I hope that helps to explain!

    14 August 2019
  • avatar

    There appears to be something wrong with my particular phone then because no matter how many times I take the train for public transit along my work route, it still frequently thinks that I take a bike instead. I have an iPhone 8. Is there anything else I could do to help the Machine Learning algorithms more? Almost every day I go into the app and confirm 10-20 different activities for that day. I love the app but just want to help it be more efficient. As a data scientist, I love the learning!

    15 August 2019
  • avatar

    As a side note, it frequently thinks I am bicycling at over 500 mph when I travel by plane :-/

    15 August 2019
  • avatar

    Hm. Sounds like we need to do some debugging then! Please send me an email from the “Feedback & Support” button inside the app. The email from there will contain debug information that can help to track down problems.

    Also, make sure you’re running the latest version of Arc (3.0.1 is the latest on the App Store), and have updated your phone to at least iOS 12.2. (iOS 12.0 and 12.1 had a bunch of problems, but they were mostly all solved by 12.2).

    15 August 2019
  • avatar

    Unfortunately, when I click on “Send a Feedback Email” nothing happens. Do I need to use the Mail app? I use Spark to manage my emails.

    I have Arc 3.0.1 and iOS 12.4.

    16 August 2019
  • avatar

    Yeah, Arc uses iOS’s built in mail sending functionality. So even if you don’t use Mail app normally, you will need your email account set up in your phone’s Settings for Arc to be able to send an email.

    16 August 2019
  • avatar

    imo this feature should be reconsidered, the ML model does not work reliably enough.

    15 July 2021