• Label them (put them in folders) by type and publication venue, for example as a journal paper, conference paper, survey paper, extended abstract, open-access, or unreferenced; some papers may have multiple labels. Quantify them by a number of citations (and remember that search tools tend to prefer highly-cited papers).
• Use the papers to map the field. Are particular institutions or authors prominent? Which venues or publication types are most significant? Is the field /t(that is, the papers are all recent), spread out over time, or some other pattern? What other features are valuable in mapping?
• Consider how these papers might be organized into a literature review. How should they be grouped? 

Using the abstracts from these papers, reconsider the research question you developed. Is it overly ambitious? Is it clear? By comparison to the existing work, where might a paper on your research question be submitted? Write down at least 5 facts, keywords, and concepts relevant to your topic.

Submit a screenshot of your folders showing your papers, your revised research question (if necessary), and the 5 facts, keywords, and concepts relevant to your topic in a document (Word or PDF). 

Topic: The effectiveness of different neural network architecture in image classification.

Research question: Which neural network architecture, if any, results in the highest accuracy when applied to a image classification task? 

Description of the topic: the effectiveness of different neural network architecture in image classification is a topic that deals with the comparison of the performance of various neural network architectures on an image classification task. The goal is to determine which architecture, if any, results in the highest accuracy. The research will involve training and testing multiple architectures such as CNN, RNN, and ResNet on a dataset and comparing their performance in terms of accuracy, speed, and robustness. This topic is a crucial one in the field of AI and ML, as it can help to improve the performance of image classification models.