hmep alternatives and similar packages
Based on the "AI" category.
Alternatively, view hmep alternatives based on common mentions on social networks and blogs.

tensorsafe
A Haskell framework to define valid deep learning models and export them to other frameworks like TensorFlow JS or Keras. 
moo
Genetic algorithm library for Haskell. Binary and continuous (realcoded) GAs. Binary GAs: binary and Gray encoding; point mutation; onepoint, twopoint, and uniform crossover. Continuous GAs: Gaussian mutation; BLXÎ±, UNDX, and SBX crossover. Selection operators: roulette, tournament, and stochastic universal sampling (SUS); with optional niching, ranking, and scaling. Replacement strategies: generational with elitism and steady state. Constrained optimization: random constrained initialization, death penalty, constrained selection without a penalty function. Multiobjective optimization: NSGAII and constrained NSGAII. 
HaVSA
HaVSA (HaveSaa) is a Haskell implementation of the Version Space Algebra Machine Learning technique described by Tessa Lau. 
Etage
A general dataflow framework featuring nondeterminism, laziness and neurological pseudoterminology.
InfluxDB  Power RealTime Data Analytics at Scale
* Code Quality Rankings and insights are calculated and provided by Lumnify.
They vary from L1 to L5 with "L5" being the highest.
Do you think we are missing an alternative of hmep or a related project?
README
Multi Expression Programming
You say, not enough Haskell machine learning libraries?
Here is yet another one!
History
There exist many other Genetic Algorithm (GA) Haskell packages. Personally I have used simple genetic algorithm, GA, and moo for quite a long time. The last package was the most preferred, but the other two are also great.
However, when I came up with this MEP paper, to my surprise there was no MEP implementation in Haskell. Soon I realized that existing GA packages are limited, and it would be more efficient to implement MEP from scratch.
That is how this package was started. I also wish to say thank you to the authors of the moo GA library, which inspired the present hmep package.
About MEP
Multi Expression Programming is a genetic programming variant encoding multiple solutions in the same chromosome. A chromosome is a computer program. Each gene is featuring code reuse.
How MEP is different from other genetic programming (GP) methods?
Consider a classical example of treebased GP.
The number of nodes to encode x^N
using a binary tree is 2N1
.
With MEP encoding, however, redundancies can be dramatically
diminished so that the
shortest chromosome
that encodes the same expression has only N/2
nodes!
That often results in significantly reduced computational costs
when evaluating MEP chromosomes. Moreover, all the intermediate
solutions such as x^(N/2)
, x^(N/4)
, etc. are provided by the
chromosome as well.
For more details, please check http://mepx.org/papers.html and https://en.wikipedia.org/wiki/Multi_expression_programming.
MEP in open source
 By Mihai Oltean, C++
 By Mark Chenoweth, Go
 Current project, Haskell
The hmep
Features
 Works out of the box. You may use one of the elaborated examples to quickly tailor to your needs.
 Flexibility. The
hmep
package provides adjustable and composable building blocks such as selection, mutation and crossover operators. One is also free to use their own operators.  Versatility.
hmep
can be applied to solve regression problems with one or multiple outputs. It means, you can approximate unknown functions or solve classification tasks. The only requirement is a custom loss function.
Getting Started
Use Stack.
$ git clone https://github.com/masterdezign/hmep.git && cd hmep
$ stack build installghc
CLI interface
A CLI interface to Haskell multi expression programming
Usage: hmep f <input file> [llength 30] [mmutation 0.05] [rvar 0.1]
[cconst 0.05] [ppopulation 200] [ttotal 200]
Available options:
h,help Show this help text
f <input file> Input file path. Format: commaseparated, two
columns.
l,length 30 Chromosome length
m,mutation 0.05 Mutation probability
r,var 0.1 Probability to generate a new variable gene
c,const 0.05 Probability to generate a new constant gene
p,population 200 Population size
t,total 200 Total number of iterations
Example: run for total of 200 algorithm iterations
$ stack exec hmep  f data/sine.txt t 200
Chromosome length: 30
Population size: 200
Mutation probability: 5.0e2
Probability to generate a new variable gene: 0.1
Probability to generate a new constant gene: 5.0e2
Probability to generate a new operator: 0.85
Reading file data/sine.txt
Fetched 50 records
Average loss in the initial population 0.6164572493880963
Population 5: average loss 0.36179141986463337
Population 10: average loss 0.35977590095295237
Population 15: average loss 0.3592976870934518
Population 20: average loss 0.35839623098861284
Population 25: average loss 0.35424451881439295
Population 30: average loss 0.31573374522629394
Population 35: average loss 0.1864152668405434
Population 40: average loss 8.966643495391169e2
Population 45: average loss 8.522968243289145e2
Population 50: average loss 8.522968243289145e2
...
Population 200: average loss 5.51041829148264e2
Interpreted expression:
v2 = x0 * x0
v4 = 0.12453785273085771 * x0
v5 = 0.12453785273085771 + v4
v6 = x0 * v4
v7 = v4 * 0.12453785273085771
v8 = v4 * v2
v12 = x0 + v5
v14 = v8 + x0
v19 = v8 * v7
v22 = v19 * v6
v23 = v22 * v12
result = v14 + v23
CLI application source is [here](app/CLI/Main.hs).
Library Example 1
Now that the package is built, run the first demo to
express cos^2(x)
through sin(x)
:
$ stack exec hmepdemo
Average loss in the initial population 15.268705681244962
Population 10: average loss 14.709728527360586
Population 20: average loss 13.497114190675477
Population 30: average loss 8.953185872653737
Population 40: average loss 8.953185872653737
Population 50: average loss 3.3219954564955856e15
Interpreted expression:
v1 = sin x0
v2 = v1 * v1
result = 1  v2
Effectively, the solution cos^2(x) = 1  sin^2(x)
was found.
Of course, MEP is a stochastic method, meaning that there is
no guarantee to find the globally optimal solution.
The unknown function approximation problem can be illustrated
by the following suboptimal solution for a given set of random
data points (blue crosses). This example was produced by another run of
the [same demo](app/Demo1/Main.hs), after 100 generations of 100 chromosomes
in each. The following expression was obtained
y(x) = 3*0.31248786462471034  sin(sin^2(x))
.
Interestingly, the approximating function lies symmetrically
inbetween the extrema of the unknown function, approximately
described by the blue crosses.
Library Example 2
A similar example is to approximate sin(x)
using only
addition and multiplication operators, i.e. with polynomials.
$ stack exec hmepsinapproximation
The algorithm is able to automatically figure out the
powers of x
. That is where MEP really shines. We [calculate](app/Demo2/Main.hs)
c'length = 30
expressions represented by each chromosome gene practically with no
additional computational penalty. We choose the best expression among those 30
in each chromosome of the population c'popSize = 200
.
In this run, we have automatically obtained a
seventh degree polynomial
coded by 14 genes. Pretty cool, huh?
v1 = 5.936286355387799e2 + 5.936286355387799e2
v4 = x0 + x0
v5 = v1 * x0
v7 = v4 * x0
v8 = v1 * v5
v9 = x0 * x0
v10 = v8 * v9
v11 = x0 * v10
v15 = 5.936286355387799e2 * x0
v18 = v10 * v11
v20 = v7 * v15
v21 = v15 + x0
v25 = v21 + v20
result = v18 + v25
Which is 0.940637136446122*x  0.118725727107756*x**3 + 0.000198691529073357*x**7
,
can be regarded as a handtuned version of x  x^3/3! + x^7/7!
(the analytic expression
is x  x^3/3! + x^5/5!  x^7/7!
).
That is impressive given that this is computed in fourteen steps!
Interestingly, we also observe that roughly half of expressions remain unused (e.g. v2, v3, v12...).
The result of approximation is [visualized](doc/sin_approx.py) below:
From the log below, one can also infer that obtained approximation is better than analytical Taylor sine expansions of 3rd and 5th orders. And naturally, is worse than the 7th order Taylor expansion:
MEP expression: Average distance for 300 points: 0.0303
3rdorder Taylor sine expansion: Average distance for 300 points: 0.3633
5rdorder Taylor sine expansion: Average distance for 300 points: 0.0688
7rdorder Taylor sine expansion: Average distance for 300 points: 0.0079
Authors
This library is written and maintained by Bogdan Penkovsky