Team:NTNU Trondheim/Modeling

From 2011.igem.org

(Difference between revisions)
(The Models)
Line 20: Line 20:
== The Models==
== The Models==
-
 
=== Systems of ODE ===
=== Systems of ODE ===
Line 27: Line 26:
=== Bayesian Hierarchy  ===
=== Bayesian Hierarchy  ===
We then wish to model the reliability for the observations... That is the probability of false positive/negative results
We then wish to model the reliability for the observations... That is the probability of false positive/negative results
-
<math>P(RTF = 1|stress) ^{T}</math>
+
P(RTF = 1|stress)<sup>T</sup>
and opposite.
and opposite.
Having the observations  from the lab  
Having the observations  from the lab  
-
"\mathtbf{x} = (x_{1} , x_{2} , \cdots , x_{n})"
+
x = (x<sub>1</sub> , x<sub>2</sub> , &middot; &middot; &middot; &middot; , x<sub>n</sub>)
-
were x_i is under condition C = 1 (stress) , and y_j is under condition C = 0 (no stress).
+
were x<sub>i</sub> is under condition C = 1 (stress) , and y<sub>j</sub> is under condition C = 0 (no stress).
=== Linear Classification ===
=== Linear Classification ===

Revision as of 20:58, 11 July 2011


Modeling

3 types of models: Systems of ODE, Bayesian hierarchy and linear classification problems (LDA or similar). To be continued....


Model Introduction

-What to model

-How to model





The Models

Systems of ODE

Bayesian Hierarchy

We then wish to model the reliability for the observations... That is the probability of false positive/negative results P(RTF = 1|stress)T and opposite. Having the observations from the lab

x = (x1 , x2 , · · · · , xn)

were xi is under condition C = 1 (stress) , and yj is under condition C = 0 (no stress).

Linear Classification

Non-linear Classification

Model Validation

References