Let Skedbee know what you did in the past, it can plan for the future.
Skedbee uses a machine learning method called Reinforcement Learning. When you start a timer to track time on a task or by adding a log in your journal, you are letting Skedbee agents know that you are working on a task and by default our agents (workers) assume that you prefer to work during that time. You have an option to update your preference and correct the agent's decision.
Skedbee uses a machine learning method called Reinforcement Learning.
Skedbee is so intelligent that, even if you forgot to start the timer and add a log in your journal a week later, Skedbee can learn and make changes to its decisions. All you have to do is, let Skedbee know when you worked and how long you worked.
To teach Skedbee, let Skedbee know when you worked and how long you worked.
Generally people's perceptions are that machine learning takes time to learn. While it's true, we have a proprietary algorithm that can learn quickly and identify changes in behaviour and adjust its learning pattern.
The reward for our agents is when you start tracking time on Skedbee suggested activities. This way, our learning agents understand that it did a better job by suggesting you the right activities at the right time and continues to improve the learning process. Skedbee's goal is to get more reward. So, it works toward suggesting you the right activity at the right time.
The first step to Skedbee's learning process is to find the probability of completing a task before it's due date given it's schedule. Skedbee agents constantly work to groom each task periodically. A small task like "get milk" or a huge task like "design a rocket booster" are considered the same initially, but prioritized differently based on time and schedule.
The second step is to use machine learning data models to find a pattern and to prioritize your tasks. Once prioritized, Skedbee weaves your tasks and preferences with time as we progress through time.
Skedbee weaves your prioritized tasks and preferences with time as we progress through time.
External dependencies (work in progress) such as weather, location, friends or group/team availability are also considered by learning agents while calculating task priority. So, using Skedbee you can predict if you are able to complete a project on time given the availability of your teammates or plan a camping trip given the weather.
Last step is to create a decision tree and to have a plan for the future. Skedbee makes a number of decisions and pre-plan a path. Decisions like, what if the suggested task is not completed in the specified hour? what if the task is 10% complete? and so on...Skedbee clearly defines a path for each action and alerts you when you are at risk of not having enough time to complete. So, Skedbee can use this decision tree to very quickly reprioritize and reschedule every task on any action you take.
Skedbee creates a decision tree and has a plan for each of your actions.
The same preferences will be used when you follow or import activities shared by others. So, activities are scheduled in the calendar right when you need them. Skedbee's goal is to make you complete your activities you added, by suggesting you at the right time with all the needed information.
Skedbee learning agents are configured to work for you. We do not use dataset from other accounts or other sites and recommend things to you. You will have the option to share your learning data models to others, if you believe others will benefit from your learning data models.
Skedbee learning agents will not use datasets from other accounts to recommend things to you.
Learning feature is configurable per activity or tasks and can be disabled anytime. Skedbee gives you full control of your account.
You may also be interested in What Is Learnable Time Blocking In Skedbee?