This article tells the story of the design of Learning by Design(tm) (LBD), a project-based inquiry approach to science learning with roots in case-based reasoning and problem-based learning, pointing out the theoretical contributions of both, classroom issues that arose upon piloting a first attempt, ways we addressed those challenges, lessons learned about promoting learning taking a project-based inquiry approach, and lessons learned about taking a theory-based approach to designing learning environments. LBD uses what we know about cognition to fashion a learning environment appropriate to deeply learning science concepts and skills and their applicability, in parallel with learning cognitive, social, learning, and communication skills. Our goal, in designing LBD, was to lay the foundation in middle school for students to be successful thinkers, learners, and decisionmakers throughout their lives and especially to help them begin to learn the science they need to know to thrive in the modern world. LB...

/pdf/problem-based-learning-meets-case-based-reasoning-in-the-21norcywah.pdf

Problem-Based Learning Meets Case-Based Reasoning in the Middle-School Science Classroom: Putting Learning by Design(tm) Into Practice

As machine learning black boxes are increasingly being deployed in domains such as healthcare and criminal justice, there is growing emphasis on building tools and techniques for explaining these black boxes in an interpretable manner. Such explanations are being leveraged by domain experts to diagnose systematic errors and underlying biases of black boxes. In this paper, we demonstrate that post hoc explanations techniques that rely on input perturbations, such as LIME and SHAP, are not reliable. Specifically, we propose a novel scaffolding technique that effectively hides the biases of any given classifier by allowing an adversarial entity to craft an arbitrary desired explanation. Our approach can be used to scaffold any biased classifier in such a way that its predictions on the input data distribution still remain biased, but the post hoc explanations of the scaffolded classifier look innocuous. Using extensive evaluation with multiple real world datasets (including COMPAS), we demonstrate how extremely biased (racist) classifiers crafted by our framework can easily fool popular explanation techniques such as LIME and SHAP into generating innocuous explanations which do not reflect the underlying biases.

https://dl.acm.org/doi/pdf/10.1145/3375627.3375830

Fooling LIME and SHAP: Adversarial Attacks on Post hoc Explanation Methods

An increasing number of decisions regarding the daily lives of human beings are being controlled by artificial intelligence and machine learning (ML) algorithms in spheres ranging from healthcare, transportation, and education to college admissions, recruitment, provision of loans, and many more realms. Since they now touch on many aspects of our lives, it is crucial to develop ML algorithms that are not only accurate but also objective and fair. Recent studies have shown that algorithmic decision making may be inherently prone to unfairness, even when there is no intention for it. This article presents an overview of the main concepts of identifying, measuring, and improving algorithmic fairness when using ML algorithms, focusing primarily on classification tasks. The article begins by discussing the causes of algorithmic bias and unfairness and the common definitions and measures for fairness. Fairness-enhancing mechanisms are then reviewed and divided into pre-process, in-process, and post-process mechanisms. A comprehensive comparison of the mechanisms is then conducted, toward a better understanding of which mechanisms should be used in different scenarios. The article ends by reviewing several emerging research sub-fields of algorithmic fairness, beyond classification.

A Review on Fairness in Machine Learning

ness of Indexes Although cases are specific, indexes to cases need to be chosen so that the case can be used in as broad a selection of situations as appropriate. Often, this approach means indexes should be more abstract than the details of a particular case. Consider, for example, a case from CHEF (Hammond 1986, 1989). CHEF just created a recipe for beef and broccoli, a stir-fried dish. When it first created the recipe and tried it out, it found that the broccoli got soggy. It fixed the order of the steps in the recipe so that the broccoli remained crisp. This case could be indexed in several ways: (1) dish is prepared by stir frying, dish includes beef, and dish includes broccoli and (2) dish is prepared by stir frying, dish includes meat, and dish includes a crisp vegetable. The first set allows this case to be recalled whenever beef and broccoli are to be stir fried together. This index, however, would not allow recall of this case, for example, when chicken and snow peas are to be stir fried. However, the order of the steps probably has to be the same as for beef and broccoli—snow peas are also a crisp vegetable that should remain crisp. Indexing by the second set of descriptors makes this case more generally applicable. Concreteness of Indexes The danger of abstract indexes is that they can be so abstract that the reasoner would never realize that a new situation had these descriptors except through extensive inference. Thus, although indexes need to be generally applicable, they need to be concrete enough so that they can be recognized with little inference. Consider another example from CHEF to illustrate this point. CHEF just created a new recipe for a strawberry souffle. It created this dish by adapting a recipe for vanilla souffle. When it first made the souffle, it fell. CHEF figured out that the problem was that the liquids and leavening were not balanced: There was too much liquid for the amount of leavening in the recipe. It also figured out that the extra liquid was because of the juice in the strawberries. It solved the problem by increasing the leavening to counter the effect of the liquid in the strawberries. This case could be indexed in several ways: (1) dish is of type souffle, and liquids and leavening are not balanced; (2) dish is of type souffle, and dish includes strawberries; (3) dish is of type souffle, and dish includes fruit; and (4) dish is of type souffle, and dish has a lot of liquid. The last three indexes are clearly better Articles

/pdf/improving-human-decision-making-through-case-based-decision-18jy1pli1o.pdf

Improving human decision making through case-based decision aiding

Case-based reasoning provides both a methodology for building systems and a cognitive model of people. It is consistent with much that psychologists have observed in the natural problem solving people do. Psychologists have also observed, however, that people have several problems in doing analogical or case-based reasoning. Although they are good at using analogs to solve new problems, they are not always good at remembering the right ones. However, computers are good at remembering. I present case-based decision aiding as a methodology for building systems in which people and machines work together to solve problems. The case-based decision-aiding system augments the person's memory by providing cases (analogs) for a person to use in solving a problem. The person does the actual decision making using these cases as guidelines. I present an overview of case-based decision aiding, some technical details about how to implement such systems, and several examples of case-based systems.

/pdf/improving-human-decision-making-through-case-based-decision-rjqgwdvt2b.pdf

A case-based reasoner can frequently benefit from using pieces of multiple previous cases in the course of solving a single problem. In our model, case pieces, called snippets, are organized around the pursuit of a goal, and there are links between the pieces that preserve the structure of reasoning. The advantages of our representational approach include: 1) The steps taken in a previous case can be followed as long as they are relevant, since the connections between steps are preserved. 2) There is easy access to all parts of previous cases, so they can be directly accessed when appropriate.

/pdf/distributed-cases-for-case-based-reasoning-facilitating-use-1dwbrgs1v3.pdf

Distributed cases for case-based reasoning; facilitating use of multiple cases

When a student is new to a domain and task he has a lot to learn to become effective. After acquiring a basic level of background knowledge, he needs to start learning how to solve problems in the domain. As a novice, doing this on his own presents significant obstacles. When he fails, assigning blame for the failure and learning from it may be impossible.
Thus, apprenticeship is valuable in helping a novice learn to carry out problem solving. We have focused on the part of apprenticeship in which the student learns by observing the expert's problem solving.
We present a model of learning by observing an expert. It assumes the existence of a competent, helpful instructor. This model shows how an interested, active student can quickly become competent in a new domain and task, through interaction with an expert.
There are several keys to the success of this learning framework. First, a student who tries to predict the expert's actions has a better understanding of his own strengths and weaknesses than a student who does not. Second, a student who attempts to draw the connections between the instructor's goals not only forms a better understanding of them, he also monitors his need to learn. The third key relates to the instructor's cooperation. The student can ask the instructor a question, and the instructor may provide hints.
Through the methods specified in the model, a student can rapidly improve his performance through exposure to a few relevant examples. Much of this early improvement is through retaining the examples as understood cases. The model shows how a student can learn without having strong knowledge of a domain and also shows how learning can benefit future learning.
The model makes contributions to the field of case-based reasoning, including the case representation, the method of case acquisition, and the methods of improving case retrieval.
The model is easily extensible to make use of new reasoning and learning methods. We propose the model as a cognitive model, and suggest testable cognitive predictions that it makes.

Learning by observing and understanding expert problem-solving

Group projects are a valuable part of the computer science curriculum. Group work can be enhanced if formation of groups is not via self-selection by the students themselves. Students who are assigned to groups are more likely to be exposed to other students with different backgrounds and abilities from which they can learn new things. However in a university with students having a wide mix of schedules, a crucial aspect of successful group formation is compatible time-schedules within a group. This paper describes a computer program designed to aid assignment to groups while helping to ensure that groups have suitable outside-of-class meeting times.

http://csis.pace.edu/~ctappert/it691-12fall/projects/groupassign-redmond.pdf

A computer program to aid assignment of student project groups

http://tavana.us/publications/ER-CLUSTER.pdf

An automated entity-relationship clustering algorithm for conceptual database design

Learning from instruction is a powerful technique for improving problem solving. An active student will predict the instructor's actions and then try to explain the differences from the predictions. We expand the concept of explanation beyond the provably correct explanations of Explanation-based learning to include other methods of explanation. The explanations can use deductions from causal domain knowledge, plausible inferences from the instructor's actions, previous cases of problem solving, and induction. The explanations result in improved diagnosis and improved future explanation. This combination of learning techniques leads to more opportunities to learn.

Combining case-based reasoning, explanation-based learning, and learning from instruction

Handle everyday research tasks with reliable, citation-backed results

Your personal Research Agent to handle research tasks with citation-backed results

Popular Tasks used by Researchers

How can I help with your research?

Meet SciSpace

Get more enhanced response by uploading the PDFs you want me to reference.

No relevant PDFs in your library

SciSpace is the AI research assistant for academics. Run systematic literature reviews on 280M+ papers, and write papers with cited sources. Trusted by 1M+ students, PhDs & researchers.

SciSpace | AI for Research

Analyze PDFs

Code & Manuscripts

Funding & Grants

Literature & Patents

Medical & Clinical Data

Systematic Review

Visualize & Present

Web & Data

Build a Google Scholar-like website for your research.

Build a website

Create charts and images for your research

Create a Chart

Write a paper for submission to a journal

Draft a manuscript

Patent Search

Design eye-catching scientific posters in minutes.

Scientific Poster Generation

Systematic Literature Review

One task is running at the moment. Your messages will be shown right after.

Drag and drop or click here to browse

Loved by <highlight>1 million+</highlight> researchers

Extract a list of specific topics and their sources from unstructured text

Topics

Compare and analyze relevant papers that matches with your search

Papers

Get insights from PDFs and bookmarked papers from your library

My library

Recent searches

Try searching for:

Catch AI-generated content in scholarly and non-scholarly content

{ai} Detector

Ai Writer

Get PDF Summaries, highlighted text explanations 

Chat with PDF

Effortlessly create in-text citations and bibliographies in APA and 2,500 other formats

Citation generator

Get explanations, summaries, and answers on academic papers

Ease up your research workflow with {scispace}'s cohort of exciting AI tools

Elevate your academic writing skills and convey your ideas the way you want

Paraphraser

Explore our range of reading and writing tools

Your file is being prepared and should be ready in a few minutes. If it's a large file, it might take a bit longer. You can close this window, and we'll email you the file when it's done.

You have reached a maximum limit of <strong>{limit}</strong> columns in the table. Remove at least <strong>1</strong> column to add or create another one.

Michael A. Redmond

Author Tools

Chat about Author