- Aug 2021
-
graciellehigino.github.io graciellehigino.github.io
-
[REFS]
Missing reference
-
(Bullock et al. 2000; Hellmann2012InfSpe?; Afkhami2014MutEff?; Godsoe et al. 2017; Siren2020IntRan?)
Check some references
-
tbl. ¿tbl:everyone?.
Fix reference to table
-
On the other hand, range preservation is also associated with ecological interactions, once connected species can be protected of climate change and invasion
I'm not sure I understand this argument
-
tbl. ¿tbl:everyone?
Fix reference to table
-
tbl. ¿tbl:everyone?
Fix reference to table
-
tbl. ¿tbl:everyone?
Fix reference to table
-
Scott2018RolHer
Check this reference
-
-
poisotlab.github.io poisotlab.github.io
-
species,
on the number of links, disconnected species, and connectance
-
fig. 5
Missing parenthesis
-
on the number of disconnected species
Is it the proportion of disconnected species that is plotted?
-
fig. 3
to remove
-
fig. 2
As we see in fig. 3
-
generality traits species
generality traits representing species
-
- Jul 2021
-
poisotlab.github.io poisotlab.github.io
-
species pairs in interactions
species pairs and interactions? Or a set of interacting species pairs
-
neural network
neural networks
-
modelling
spelled "modelling" and "modeling" in the text
-
our understanding
to further our understanding of what, exactly?
-
embedding
remove the second embedding
-
imputed usin
the imputed network using
-
are, further,
too many comas
-
the structure of
I would remove "the structure of" in this last sentence since our aim is broader than that.
-
concepts
concepts in
-
approaches
approach
-
contest
Are we really contesting that?
-
predictions of binary interactions can be more readily interpreted.
Repeated sentence
-
-
poisotlab.github.io poisotlab.github.io
-
First
It would be interesting to know the proportion of interacting species pairs in Europe that were predicted by the model (a bit like what you did in the following paragraphs).
-
though
through
-
ρ
I am not sure I understand this inequality. How can a matrix be >= to a real number?
I'm guessing you meant that each element of the matrix = 1 if the corresponding element in multiplied matrices is more than p, and 0 otherwise? I think we should make it more explicit and mathematically correct.
-
L^R^′≥ρ\hat{\mathcal{R}}' \ge \rhoR^
Shouldn't it be lr (small caps)?
-
sub-spaces
subspaces
Same below.
-
eigenvalues
These are the singular values, right? Or eigenvalues?
-
rank
If I'm not mistaken, rank and dimensions are used interchangeably? Maybe we should use consistent terminology.
-
*
Why is there a star here?
-
XXX species
Were there any species in the mammalian European metaweb not connected to any other species? Since we made sure all species had at least one interaction in the Canadian metaweb, I think it would be important to know if this is also the case in the mammalian European subgraph. If this was the case, did you keep them in the network?
-
European metaweb
Maybe describe a bit more what the European metaweb is (e.g. species included, types of interactions).
-
- Jun 2021
-
poisotlab.github.io poisotlab.github.io
-
allowing
"allowing" is written twice in the same sentence
-
Where,
Remove this word.
-
A promising application of machine learning in natural sciences is Scientific-Machine Learning (SciML), a framework that combines machine learning with mechanistic models
I'm wondering if we should remove this sentence. I feel it doesn't flow very well and we don't really talk about SciML later in the text.
-
interaction
Just to be sure, do these scores represent in fact the probabililty of cooccurrence?
-
neural-network
neural network
-
for prediction
for the prediction
-
show a
where we show how a simple neural network OR where we show that a simple neural network?
-
- Feb 2021
-
poisotlab.github.io poisotlab.github.io
-
This could potentially be solved through our framework of predicting networks first, interactions next, and finally the realized species pool.
I think this might be confusing, as readers could think this is the actual framework of our paper. We could maybe precise that we could use our framework the other way around, by using our predictions of ecological networks and species interactions to make better predictions of species pools.
-
all interactions occur between species in each pool
I think this definition of bipartite network is a little confusing. We could say something like:
"Bipartite networks are divided into two disjoint sets of species and interactions occur between members of different sets (e.g. plant-pollinator and host-parasite networks)"
-
- Jan 2020
-
poisotlab.github.io poisotlab.github.io
-
s and
Hyperparameters p and phi?
-