Paulo Levi
@pauloleve.bsky.social
Residente de Psiquiatria - ESP Sobral
M.D, resident in psychiatry.
M.D, resident in psychiatry.
Risperidone does not induce mania. However it may induce a state of restless agitation called akathisia.
December 29, 2024 at 6:53 PM
Risperidone does not induce mania. However it may induce a state of restless agitation called akathisia.
Do you see any patterns in these groups? I honestly don't haha, but it was a fun journey. To me, the main insight is that dreams sound as hermetic as ever, even with new technologies.
November 20, 2024 at 5:01 PM
Do you see any patterns in these groups? I honestly don't haha, but it was a fun journey. To me, the main insight is that dreams sound as hermetic as ever, even with new technologies.
Here you can see a common partially unprocessed dream, with colors telling each word group. (grey words were either stopwords or just were not frequent enough to get included in the analysis)
November 20, 2024 at 5:01 PM
Here you can see a common partially unprocessed dream, with colors telling each word group. (grey words were either stopwords or just were not frequent enough to get included in the analysis)
I've found three main topics and created word clouds given the frequency of each one in the text. Which looks more or less like this
November 20, 2024 at 5:00 PM
I've found three main topics and created word clouds given the frequency of each one in the text. Which looks more or less like this
Why go through all that work? because now with this map, I'm able to create k-clusters of meaning (or topics) that are frequent in our sample dreams.
November 20, 2024 at 4:59 PM
Why go through all that work? because now with this map, I'm able to create k-clusters of meaning (or topics) that are frequent in our sample dreams.
After pre-processing, I used Latent semantic analysis(LSA) to create a virtual space of semantic similarity. In LSA, words that co-occur probably have more semantic similarity. Words such as "puppy" would be closer to "dog" than "puppet". Creating a "map" of distances between each of these words.
November 20, 2024 at 4:56 PM
After pre-processing, I used Latent semantic analysis(LSA) to create a virtual space of semantic similarity. In LSA, words that co-occur probably have more semantic similarity. Words such as "puppy" would be closer to "dog" than "puppet". Creating a "map" of distances between each of these words.
So, after collecting roughly all dreams available at that site words would go through a pre-processing stage: Stopwords (such as 'the, what, why') would be removed, words were tokenized and words would be lemmatized, so it would look like this
November 20, 2024 at 4:53 PM
So, after collecting roughly all dreams available at that site words would go through a pre-processing stage: Stopwords (such as 'the, what, why') would be removed, words were tokenized and words would be lemmatized, so it would look like this
Now to quantify, measure, and analyze dreams seems to be an extraordinary task and perhaps even a profane one: to use cold numbers to discern the indiscernible. Lucky for us, Natural Language Processing attempts to do exactly that in this Promethean fight.
November 20, 2024 at 4:50 PM
Now to quantify, measure, and analyze dreams seems to be an extraordinary task and perhaps even a profane one: to use cold numbers to discern the indiscernible. Lucky for us, Natural Language Processing attempts to do exactly that in this Promethean fight.
To first understand what is going on with our dreams, we need data. As I'm currently unable to go through the process of recruiting patients for that (given time and energy constraints), I have used dreams from an online database from various research studies (available at dreambank.net)
DreamBank
dreambank.net
November 20, 2024 at 4:49 PM
To first understand what is going on with our dreams, we need data. As I'm currently unable to go through the process of recruiting patients for that (given time and energy constraints), I have used dreams from an online database from various research studies (available at dreambank.net)
Reposted by Paulo Levi
I did a version about psychiatry papers back then 😆
November 18, 2024 at 8:38 PM
I did a version about psychiatry papers back then 😆
After pre-processing, I used Latent semantic analysis(LSA) to create a virtual space of semantic similarity. In LSA, words that co-occur have more semantic similarity. ( "puppy" would be closer to "dog" than "puppet"). With this, I created a "map" of the distances in-between each of these words.
November 19, 2024 at 9:44 PM
After pre-processing, I used Latent semantic analysis(LSA) to create a virtual space of semantic similarity. In LSA, words that co-occur have more semantic similarity. ( "puppy" would be closer to "dog" than "puppet"). With this, I created a "map" of the distances in-between each of these words.
So, after collecting roughly all dreams available at that site words would go through a pre-processing stage: Stopwords (such as 'the, what, why') would be removed, words were tokenized and words would be lemmatized, so it would look like this
November 19, 2024 at 9:40 PM
So, after collecting roughly all dreams available at that site words would go through a pre-processing stage: Stopwords (such as 'the, what, why') would be removed, words were tokenized and words would be lemmatized, so it would look like this
Now to quantify, measure and analyze dreams seems to be an extraordinary task and perhaps even a profane one: to use cold numbers to discern the indiscernible. Lucky for us, Natural Language Processing attempts to do exactly that in this Promethean fight.
November 19, 2024 at 9:37 PM
Now to quantify, measure and analyze dreams seems to be an extraordinary task and perhaps even a profane one: to use cold numbers to discern the indiscernible. Lucky for us, Natural Language Processing attempts to do exactly that in this Promethean fight.
To first understand what is going on with our dreams, we need data. As I'm currently unable to go through the process of recruiting patients for that (given time and energy constraints), I have used dreams from an online database from various research studies (available at dreambank.net)
DreamBank
dreambank.net
November 19, 2024 at 9:36 PM
To first understand what is going on with our dreams, we need data. As I'm currently unable to go through the process of recruiting patients for that (given time and energy constraints), I have used dreams from an online database from various research studies (available at dreambank.net)
Unfortunately, due to the scarcity of primary literature, we were unable to compare MPH and ATX groups directly.
October 10, 2024 at 7:14 PM
Unfortunately, due to the scarcity of primary literature, we were unable to compare MPH and ATX groups directly.
📍Given bias considerations, we have a significant, but very low certainty that Methylphenidate (MPH) improves Inattention, hyperactivity and irritability symptoms.
In MPH vs PLA comparison, there was a high effect size increase in dropout due to adverse effects.
In MPH vs PLA comparison, there was a high effect size increase in dropout due to adverse effects.
October 10, 2024 at 7:12 PM
📍Given bias considerations, we have a significant, but very low certainty that Methylphenidate (MPH) improves Inattention, hyperactivity and irritability symptoms.
In MPH vs PLA comparison, there was a high effect size increase in dropout due to adverse effects.
In MPH vs PLA comparison, there was a high effect size increase in dropout due to adverse effects.
📍 We have moderate certainty of evidence that atomoxetine (ATX) improves Inattention and hyperactivity symptoms with a moderate to high effect size when compared to Placebo (PLA) groups.
In ATX vs PLA, there was no significant difference in dropout due to adverse effects.
In ATX vs PLA, there was no significant difference in dropout due to adverse effects.
October 10, 2024 at 7:01 PM
📍 We have moderate certainty of evidence that atomoxetine (ATX) improves Inattention and hyperactivity symptoms with a moderate to high effect size when compared to Placebo (PLA) groups.
In ATX vs PLA, there was no significant difference in dropout due to adverse effects.
In ATX vs PLA, there was no significant difference in dropout due to adverse effects.
Recomendo o som das ondas e Neve de primavera, os livros subsequentes da saga nem tanto. Vc tá lendo oq dele?
October 1, 2024 at 10:09 AM
Recomendo o som das ondas e Neve de primavera, os livros subsequentes da saga nem tanto. Vc tá lendo oq dele?
Acho que só rola se Michael kirkbride voltasse a ter influência na série. A parte técnica nunca foi o forte da Bethesda, mas ela sempre se destacou pela questão narrativa e world building imersivo.
September 20, 2024 at 2:40 PM
Acho que só rola se Michael kirkbride voltasse a ter influência na série. A parte técnica nunca foi o forte da Bethesda, mas ela sempre se destacou pela questão narrativa e world building imersivo.
não só isso, as vezes eles dão informação errada sobre como chegar lá. Mas faz parte da imersão do jogo, aí vc passa um tempo pensando em craftar um feitiço que faça vc ficar rápido pra caralho e isso não ser um problema.
September 11, 2024 at 6:21 PM
não só isso, as vezes eles dão informação errada sobre como chegar lá. Mas faz parte da imersão do jogo, aí vc passa um tempo pensando em craftar um feitiço que faça vc ficar rápido pra caralho e isso não ser um problema.
Morrowind já tinha, não sei se já tinha em daggerfall também.
Recomendo fortemente Morrowind, lore fantastica, jogabilidade é um pouco datada mas quem gosta de jogar como spellcaster fica massa por esses motivos.
Recomendo fortemente Morrowind, lore fantastica, jogabilidade é um pouco datada mas quem gosta de jogar como spellcaster fica massa por esses motivos.
September 11, 2024 at 6:13 PM
Morrowind já tinha, não sei se já tinha em daggerfall também.
Recomendo fortemente Morrowind, lore fantastica, jogabilidade é um pouco datada mas quem gosta de jogar como spellcaster fica massa por esses motivos.
Recomendo fortemente Morrowind, lore fantastica, jogabilidade é um pouco datada mas quem gosta de jogar como spellcaster fica massa por esses motivos.