@@ -53,29 +53,29 @@ code.
5353In some cases we would like to know the causal effect of some intervention but we do not
5454have the counterfactual, making conventional methods of statistical analysis infeasible.
5555However, it may still be possible to get an unbiased estimate of the causal effect (ATE,
56- ATE, or ITT) by predicting the counterfactual and comparing it to the observed outcomes.
57- This is the approach CausalELM takes to conduct interrupted time series analysis,
58- G-Computation, double machine learning, and metalearning via S-Learners, T-Learners,
59- X-Learners, R-learners, and doubly robust estimation. In interrupted time series analysis,
60- we want to estimate the effect of some intervention on the outcome of a single unit that we
61- observe during multiple time periods. For example, we might want to know how the
62- announcement of a merger affected the price of Stock A. To do this, we need to know what the
63- price of stock A would have been if the merger had not been announced, which we can predict
64- with machine learning methods. Then, we can compare this predicted counterfactual to the
65- observed price data to estimate the effect of the merger announcement. In another case, we
66- might want to know the effect of medicine X on disease Y but the administration of X was not
67- random and it might have also been administered at mulitiple time periods, which would
68- produce biased estimates. To overcome this, G-computation models the observed data, uses the
69- model to predict the outcomes if all patients recieved the treatment, and compares it to the
70- predictions of the outcomes if none of the patients recieved the treatment. Double machine
71- learning (DML) takes a similar approach but also models the treatment mechanism and uses it
72- to adjust the initial estimates. This approach has three advantages. First, it is more
73- efficient with high dimensional data than conventional methods. Metalearners take a similar
74- approach to estimate the CATE. While all of these models are different, they have one thing
75- in common: how well they perform depends on the underlying model they fit to the data. To
76- that end, CausalELMs use bagged ensembles of extreme learning machines because they are
77- simple yet flexible enough to be universal function approximators with lower varaince than
78- single extreme learning machines.
56+ ATE, or ITT) by predicting the counterfactual and comparing it to the observed outcomes if
57+ we are able to observe all potential confounders. This is the approach CausalELM takes to
58+ conduct interrupted time series analysis, G-Computation, double machine learning, and
59+ metalearning via S-Learners, T-Learners, X-Learners, R-learners, and doubly robust
60+ estimation. In interrupted time series analysis, we want to estimate the effect of some
61+ intervention on the outcome of a single unit that we observe during multiple time periods.
62+ For example, we might want to know how the announcement of a merger affected the price of
63+ Stock A. To do this, we need to know what the price of stock A would have been if the merger
64+ had not been announced, which we can predict with machine learning methods. Then, we can
65+ compare this predicted counterfactual to the observed price data to estimate the effect of
66+ the merger announcement. In another case, we might want to know the effect of medicine X on
67+ disease Y but the administration of X was not random and it might have also been
68+ administered at mulitiple time periods, which would produce biased estimates. To overcome
69+ this, G-computation models the observed data, uses the model to predict the outcomes if all
70+ patients recieved the treatment, and compares it to the predictions of the outcomes if none
71+ of the patients recieved the treatment. Double machine learning (DML) takes a similar
72+ approach but also models the treatment mechanism and uses it to adjust the initial estimates.
73+ This approach has three advantages. First, it is more efficient with high dimensional data
74+ than conventional methods. Metalearners take a similar approach to estimate the CATE. While
75+ all of these models are different, they have one thing in common: how well they perform
76+ depends on the underlying model they fit to the data. To that end, CausalELMs use bagged
77+ ensembles of extreme learning machines because they are simple yet flexible enough to be
78+ universal function approximators with lower varaince than single extreme learning machines.
7979</p >
8080
8181<h2 >CausalELM Features</h2 >
@@ -92,20 +92,20 @@ single extreme learning machines.
9292
9393<h2 >What's New?</h2 >
9494<ul >
95+ <li >See the JuliaCon 2024 CausalELM demonstration <a href =" https://www.youtube.com/watch?v=hh_cyj8feu8&t=26s " >here.
96+ <li >Model summaries include confidence intervals and marginal effects<li >
9597 <li >Now includes doubly robust estimator for CATE estimation</li >
9698 <li >All estimators now implement bagging to reduce predictive performance and reduce variance</li >
9799 <li >Counterfactual consistency validation simulates more realistic violations of the counterfactual consistency assumption</li >
98100 <li >Uses a simple heuristic to choose the number of neurons, which reduces training time and still works well in practice</li >
99101 <li >Probability clipping for classifier predictions and residuals is no longer necessary due to the bagging procedure</li >
100- <li >CausalELM talk has been accepted to JuliaCon 2024!</li >
101102</ul >
102103
103104<h2 >What's Next?</h2 >
104105<p >
105106Newer versions of CausalELM will hopefully support using GPUs and provide interpretations of
106- the results of calling validate on a model that has been estimated. In addition, some
107- estimators will also support using instrumental variables. However, these priorities could
108- also change depending on feedback recieved at JuliaCon.
107+ the results of calling validate on a model that has been estimated. We may also add some other
108+ features depending on demand and feedback.
109109</p >
110110
111111<h2 >Disclaimer</h2 >
0 commit comments