xuhongzuo / deep-iforest Goto Github PK

hello , firstable i wanna thank you for ur amazing work
so i have feature vectors extracted from videos and i wanna use ur algorithm to detect anomalies
my question is , should i modify smthing , cause normally the model takes in tabular , graph or ts form of data

I can't install torch-scatter==2.0.7 because it tells me that only 2.0.4 and 2.0.5 are supported for torch==1.5.0+cu101, I wonder how can I fix this problem?

'''
pip install --verbose --no-index torch-scatter==2.0.7 -f https://pytorch-geometric.com/whl/torch-1.5.0+cu101.html
ERROR: Could not find a version that satisfies the requirement torch-scatter==2.0.7 (from versions: 2.0.4, 2.0.5)
ERROR: No matching distribution found for torch-scatter==2.0.7
'''

also the same problem for torch-sparse

'''
ERROR: Could not find a version that satisfies the requirement torch-sparse==0.6.9 (from versions: 0.6.2, 0.6.3, 0.6.4, 0.6.5, 0.6.6, 0.6.7)
ERROR: No matching distribution found for torch-sparse==0.6.9
'''

Thanks for your work.
The UCR natural dataset in your repository is a univariate dataset. Does DIF also support multi-variate time series dataset?

@xuhongzuo. Hello!
I came here when I searched for Isolation Forest Algorithms. Do you have any list of homepages where you showcase your academic work? Just in case I missed any awesome work.

您好，麻烦问一个问题：在进行实时时序异常检测的时候，使用您的算法得到了异常得分，怎么判断是否有异常？是使用阈值吗？如果使用阈值的话该如何设置？非常感谢。