Exam AWS Certified Machine Learning - Specialty topic 1 question 218 discussion

C. Train an Amazon SageMaker Neural Topic Model (NTM) model to generate the product categories. The task is to build a machine learning model to categorize documents for all the company's products. Among the given options, training an Amazon SageMaker Neural Topic Model (NTM) model would be the most efficient and effective solution. An NTM model can identify topics in text data and group similar documents into specific categories, making it a suitable model for document categorization. With an NTM model, the data scientist would not need to define product categories beforehand, as the model would automatically group similar documents into topics. This saves time and resources compared to the other options.

A. NO - no need to build a custom model B. NO - k-means is supervised model C. YES - unsupervised clustering algorithm D. NO - Blazing Text will do word embedding, not classification

C. Train an Amazon SageMaker Neural Topic Model (NTM) model to generate the product categories. -- option doesn't talk about any classification activity

Neural Topic Model (NTM) is one of the built-in algorithms of Amazon SageMaker that can perform topic modeling on text data. Topic modeling is a technique that can discover latent topics or themes from a collection of documents. Topic modeling can be used for document categorization by assigning each document to one or more topics based on its content.

Blazing Text is only for supervised problems.

Assign pre-defined categories to documents in a corpus: categorize books in a library into academic disciplines - BlazingText algorithm https://docs.aws.amazon.com/sagemaker/latest/dg/algos.html

Option C is wrong because it suggests using a Neural Topic Model (NTM) to categorize documents. While NTM can be used to discover the underlying topics in a corpus of documents, it may not be the most efficient solution for categorizing documents for specific products. NTM is more suited for unsupervised learning problems where the goal is to discover the underlying themes or topics of the document corpus. In this scenario, the data scientist needs to categorize documents based on predefined product categories. Therefore, a supervised learning algorithm like a text classification model would be more suitable. Amazon SageMaker Blazing Text algorithm provides an efficient and scalable solution for text classification problems.

No predefined product category: topic modeling with NTM or LDA (Organize a set of documents into topics (not known in advance): tag a document as belonging to a medical category based on the terms used in the document.) Predefined product category: topic modeling with blazing text (categorize books in a library into academic disciplines) c