I am given a data set where I have a sequence of tasks related to a particular person:
我得到一个数据集,我有一系列与特定人相关的任务:
Person Timestamp Task
1 0000 3
1 0001 6
1 0002 1
2 0000 9
2 0001 8
These tasks are usually performed in order.
这些任务通常按顺序执行。
How can we predict the next 10 tasks performed by each person?
我们如何预测每个人执行的下10个任务?
1 个解决方案
#1
1
I believe you want to predict the future event of each person based on the history of the person. For that one event of a person cannot be used. You must have a history of his past events which can then be use to make predictions. Kindly refer to enter link description here. What you are trying to achieve is indeed an interesting task but the data may be insufficient. For example lets take person with id 1. In that case this is purely a time series forecasting problem. LSTM's could be used learn from the history of the person then predict.
我相信你想根据这个人的历史来预测每个人的未来事件。因为一个人的事件不能被使用。您必须拥有过去事件的历史记录,然后才能用于预测。请参阅此处输入链接说明。你想要实现的目标确实是一项有趣的任务,但数据可能不足。例如,让我们采用id为1的人。在这种情况下,这纯粹是一个时间序列预测问题。可以使用LSTM从人的历史中学习然后预测。
#1
1
I believe you want to predict the future event of each person based on the history of the person. For that one event of a person cannot be used. You must have a history of his past events which can then be use to make predictions. Kindly refer to enter link description here. What you are trying to achieve is indeed an interesting task but the data may be insufficient. For example lets take person with id 1. In that case this is purely a time series forecasting problem. LSTM's could be used learn from the history of the person then predict.
我相信你想根据这个人的历史来预测每个人的未来事件。因为一个人的事件不能被使用。您必须拥有过去事件的历史记录,然后才能用于预测。请参阅此处输入链接说明。你想要实现的目标确实是一项有趣的任务,但数据可能不足。例如,让我们采用id为1的人。在这种情况下,这纯粹是一个时间序列预测问题。可以使用LSTM从人的历史中学习然后预测。