Reducing calculus trauma, and teaching AI to smell
Aug 31, 2023
auto_awesome
Laird Kramer, professor of physics at FIU, discusses improving calculus instruction and reducing calculus trauma. Science staffers share their own calculus struggles. Emily Mayhew, professor at Michigan State University, talks about teaching AI to predict smells based on chemical structure.
Improving calculus instruction in universities through active learning can increase student success and interest in STEM fields.
Machine learning can be used to predict human perception of odors based on chemical structure, opening doors for advancements in designing smells.
Deep dives
Improving Calculus Instruction for STEM Fields
Physicist and Education Researcher, Laird Kramer, discusses his work revealing how improving calculus instruction at universities might encourage more students to stick with STEM fields. The study found that active learning, such as solving problems in small groups and engaging in discussions, helps students better understand and appreciate calculus, leading to higher success rates and increased interest in STEM subjects.
Digitizing the Sense of Smell
Researcher Emily Mayhew discusses the challenges of digitizing the sense of smell and the potential applications of machine learning in mapping odor molecules to human perception. By training an AI model using a large dataset of odor descriptors, the study successfully predicted how humans would perceive the smell of new molecules based solely on their chemical structure. This research opens doors for faster prediction of odor perceptions and understanding the relationship between molecules and odors.
The Impact of Improved Calculus Instruction
Inadequate calculus instruction has been a major factor in student dropout rates from STEM fields. The study by Laird Kramer found that improving calculus teaching methods not only increased student success in calculus courses but also led to better overall academic performance. The treatment group showed a 0.4 increase in GPA and a 11% higher passing rate in calculus compared to the control group. These findings emphasize the importance of effective teaching methods in retaining students in STEM programs.
Mapping Smells to Molecules Using AI
Researchers used a large dataset and machine learning to predict how different molecules would smell to humans. By training an AI model on odor descriptors associated with chemical structures, the study achieved impressive accuracy in predicting human perception of novel smell molecules that had not been smelled before. The results suggest the potential for future advancements in designing smells and the digitization of the sense of smell.
How active learning improves calculus teaching, and using machine learning to map odors in the smell space
First up on this week’s show, Laird Kramer, a professor of physics and faculty in the STEM Transformation Institute at Florida International University (FIU), talks with host Sarah Crespi about students leaving STEM fields because of calculus and his research into improving instruction.
We also hear from some Science staffers about their own calculus trauma, from fear of spinning shapes to thinking twice about majoring in physics (featuring Kevin McLean, Paul Voosen, Lizzie Wade, Meagan Cantwell, and FIU student and learning assistant Carolyn Marquez).
Next on the show, can a computer predict what something will smell like to a person by looking at its chemical structure? Emily Mayhew, a professor in the department of food science and human nutrition at Michigan State University, talks about how this was accomplished using a panel of trained smellers, and what the next steps are for digitizing the sense of smell.
This week’s episode was produced with help from Podigy.