CLaSMO: Conditional Latent Space Scaffold Optimization

CLaSMO is a web-application that allows modifications of input chemical compounds based on the target property.

bayesmedaug: Bayesian Optimization Library for Medical Image Segmentation, Python Library.

bayesmedaug optimizes your data augmentation hyperparameters for medical image segmentation tasks by using Bayesian Optimization.

import torch
import bayesmedaug
from bayesmedaug import VanillaUNet, Trainer, BinaryListed, BOMed
from bayesmedaug import Rotate, ZoomOut

auglist = [Rotate]
params = {'angle': (0.2,2.8),'p_'+ZoomOut.__name__: (0, 1)}                                      
trainer = Trainer(
    model = VanillaUNet,...
)
optimizer = BOMed(
    f = trainer.train,
    pbounds = params,
    random_state = 1,
)
optimizer.maximize(
    init_points = 15,
    n_iter = 15,
)

VarRedOpt: A Framework for Variance Reduction
An R Library published on CRAN for efficient Monte Carlo Simulation techniques with built-in simulation functions for Option Pricing. 

install.packages("VarRedOpt")
library(VarRedOpt)

# Simulate Asian Options with Antithetic Variates
sim.outer(n=1e5, d=3, q.outer = sim.AV, 
               q.av = myq_asian, K = 100, ti = (1:3)/12, r = 0.03, sigma = 0.3, S0 = 100)
#>    Estimation StandardError 
#>    4.55000000    0.02285073

Recompy: A Python Library for Recommender Systems

Recompy is a library for recommender systems. It provides an easy framework to train different models, calculate similarities and recommend items for both existing and new users.

Test Recompy and get movie recommendations from here!

from recompy import load_movie_data, FunkSVD

# get MovieLens data
data = load_movie_data()
# initialization of FunkSVD model
myFunk = FunkSVD()
# training of the model
myFunk.fit(data)

# Create new user. Key:Item ID, Value:Rating
new_user = {'1':5,
            '2':4,
            '4':3}
            
# To find the most similar user resulting from cosine similarity. Recommend 5 items using the most similar user 
myFunk.get_recommendation_for_new_user(new_user, similarity_measure = 'cosine_similarity', 
                                       howManyUsers = 1, howManyItems = 5)