Cognitive models

# Cognitive models
### <a href='https://cdsbasel.github.io/neuralnetworks/'> Introduction to Neural Networks </a> <a href='https://cdsbasel.github.io/neuralnetworks/menu/materials.html'> </a>  <a href='https://cdsbasel.github.io/dataanalytics/'> </a>  <a href='mailto:dirk.wulff@unibas.ch'> 
### May 2020

---

<div class="my-footer">
 
 
 <img src="https://raw.githubusercontent.com/therbootcamp/therbootcamp.github.io/master/_sessions/_image/by-sa.png" height=14 style="vertical-align: middle"/>
 
 <a href="https://cdsbasel.github.io/neuralnetworks/">
 
 
 cdsbasel.github.io/neuralnetworks/
 
 
 </a>
 <a href="https://cdsbasel.github.io/neuralnetworks/">
 
 Introduction to Neural Networks | May 2020
 
 </a>
 
 </div>

---

# Neural processing in the brain

<ul>
 <li class="m1">Efforts are made to build neural network models that <high>mimick the brain's architectural properties</high> concerning, e.g., the visual system.</li>
</ul>

]

<img src="image/brain.png"> 
from <a href="https://neuwritesd.org/2015/10/22/deep-neural-networks-help-us-read-your-mind/">neuwritesd.org</a>, see <a href="https://www.pnas.org/content/111/23/8619">this</a>

]

---

# Semantic space

<ul>
 <li class="m1">Many cognitive models build on some for of a <high>mnemonic representation of words</high> and their meaning, which can be apprximated using network-based vector space models.</li>
 <li class="m2"><high>Tasks</high> linked to vector space models:</li>
 <ul>
 <li>Verbal fluency</li>
 <li>Similarity ratings</li>
 <li>Paired associated learning</li>
 <li>Free recall</li>
 <li>Verbal reasoning</li>
 <li>etc.</li>
 </ul>
</ul>

]

]

---

# Walks in semantic space

<ul>
 <li class="m1">A popular idea is that retrieval from (semantic) memory acts like a <high>random walk through the word vector space</high>.</li>
 <li class="m2"><high>Verbal fluency</high> sequences are consistent with this notion barring the occasional jump.</li>
</ul>

]

]

---

# Walks in semantic space

]

]

---

# Walks in semantic space

]

]

---

# Walks in semantic space

]

]

---

# Hopfield

<ul>
 <li class="m1">Hopfield networks are a form of recurrent neural network, that can be used as <high>models of recognition</high>.</li>
 <li class="m2">They can be using Hebbian learning: <high>fire together, wire together</high>.</li>
</ul>

`$$\Large \Delta w_{ij}=\eta \cdot a_{i}\cdot a_{j}$$`

]

]

---

# Hopfield

`$$\Large \Delta w_{ij}=\eta \cdot a_{i}\cdot a_{j}$$`

]

]

---

# Hopfield

`$$\Large \Delta w_{ij}=\eta \cdot a_{i}\cdot a_{j}$$`

]

]

---

# Hopfield

<ul>
 <li class="m1">Hopfield networks are a form of recurrent neural network, that can be used as <high>models of recognition and associative memory</high> .</li>
 <li class="m2">They can be using Hebbian learning: <high>fire together, wire together</high>.</li>
</ul>

`$$\Large \Delta w_{ij}=\eta \cdot a_{i}\cdot a_{j}$$`

]

]

---

# Cognitive consistency

<ul>
 <li class="m1">Recursive networks can be used to predict the dynamics and outcomes of <high>consistency-driven cognitive reasoning</high>.</li>
 <li class="m2">Domains:</li>
 <ul>
 <li>Dissonance theory</li> 
 <li>Opinion formation</li> 
 <li>Inference</li>
 <li>Decision making</li> 
 </ul>
</ul>

]

<img src="image/fest1.png" height=460px> 
from <a href="https://journals.sagepub.com/doi/10.1207/s15327957pspr0101_3">Read, Vanman, & Miller (1997)</a>

]

---

# Cognitive consistency

]

<img src="image/fest2.png" height=460px> 
from <a href="https://journals.sagepub.com/doi/10.1207/s15327957pspr0101_3">Read, Vanman, & Miller (1997)</a>

]

---

# Cognitive consistency

]

<img src="image/socpsy1.png" height=460px> 
from <a href="https://journals.sagepub.com/doi/10.1207/s15327957pspr0101_3">Read, Vanman, & Miller (1997)</a>

]

---

# Cognitive consistency

]

]

---

# Cognitive consistency

]

<img src="image/glöck.png" height=460px> 
from <a href="https://www.econstor.eu/bitstream/10419/26939/1/572287593.PDF">Glöckner & Betsch (2008)</a>

]

---

# Cognitive consistency

]

<img src="image/glöck2.png" height=460px> 
from <a href="https://www.econstor.eu/bitstream/10419/32204/1/605761744.pdf">Glöckner & Herbold(2008)</a>

]

---

<h1><a href="https://cdsbasel.github.io/neuralnetworks/menu/schedule & materials.html">Materials</a></h1>