Introducing 

Prezi AI.

Your new presentation assistant.

Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.

Loading content…
Loading…
Transcript

PROSPECT THEORY: AN ANALYSIS OF DECISION UNDER RISK

Daniel Kahneman and Amos Tversky

Miguel Morales

Júlia Garcia

Neil de la Fuente

Laia Vilardell

14/ 12/ 2021

UTILITY THEORY

UTILITY THEORY

Utility function: calculation for how much someone desires something.

Utility theory states that costumers when facing a purchase decision, rank the products in their minds based on the highest degree of satisfaction.

EXAMPLE

Mario: 45 years old, hired by a new company. Enjoys luxurious gadgets and is rich enough to posses those devices.

A new smartphone is presented to the VIP clients (costs $1,100). Mario prefers this new version.

For his satisfaction (utility), owning this new device makes him feel important and appreciated.

PROSPECT THEORY

Prospect theory: people value certainty over risk, particularly if gains and losses are equal in likelihood

Value function:

S-shaped value curve that is:

concave for gains and convex (and steeper) for losses.

PROSPECT THEORY

Key concepts:

- Risk aversion for sure gains

- Risk seeking for sure losses

- Fear of losing > joy of gaining

Weighting function:

- Overweight low-probability events.

- Underweight highly certain outcomes.

Results of the experiment

Principal sections and Experiment

Certainty, Probability, and Possibility

Certainty, Probability, and Possibility

The Isolation Effect

The Isolation Effect

The Weighting Function

The link below takes us to a video that demonstrates the Weighting Function

The Weighting Function

https://youtu.be/lhA6ZPwKyUI

EXAMPLE

We will present a model that tells us if our customers are going to default on their payments when we extend credit to them. We will use this model to decide if we give him the credit or not.

APPLICATION TO AI

Model design

We need to be very conservative doing predictions, because we need to make sure that the customer is going to pay with a very high probability.

Model design

Unbalanced sample

Techniques

Because the majority of people have paid correctly, when we train our model, it will tend to predict that customers will pay correctly. That is because as we have seen in the historical sample we have an unbalanced sample.

So, we have to train our model in a very conservative way. For it we can apply different techniques :

Conclusion

So as we have seen in prospect theory, here in the model we apply techniques related to risk aversion to earn money. Because the bank wants to ensure that they will earn the money from the interest.

Conclusion

CONCLUSIONS

Amos and Kahneman were game-changers in terms of understanding the way we make decisions.

CONCLUSIONS

This was their masterpiece, where they refuted the utility theory by this more accurate and realistic prospect theory, explaining our way to decide under risk.

Questions?

TIME FOR QUESTIONS!

THANK YOU!

Prospect Theory:

An analysis of decision under risk

Learn more about creating dynamic, engaging presentations with Prezi