Identifying alphabet sequences with Conditional Random Fields
Conditional Random Fields (CRFs) are probabilistic models that are frequently used to analyze structured data. We use them to label and segment sequential data in various forms. Following are some of the most common use cases where CRFs are applied:
- Handwriting recognition
- Character recognition
- Object detection
- Named entity recognition
- Gene prediction
- Image segmentation
- Part of speech tagging
- Noise reduction
One item of note regarding CRFs is that they are discriminative models. Contrast this with HMMs, which are generative models.
We can define a conditional probability distribution over a labeled sequence of measurements. We will use this to build a CRF model. In HMMs, we define a joint distribution over the observation sequence and the labels.
One of the main advantages of CRFs is that they are conditional by nature. This is not the case with HMMs. CRFs do not assume...
