I have a set of 20 small document which talks about a particular kind of issue (training data). Now i want to identify those docs out of 10K documents, which are talking about the same issue.
For the purpose i am using the doc2vec implementation:
from gensim.models.doc2vec import Doc2Vec, TaggedDocument
from nltk.tokenize import word_tokenize
# Tokenize_and_stem is creating the tokens and stemming and returning the list
# documents_prb store the list of 20 docs
tagged_data = [TaggedDocument(words=tokenize_and_stem(_d.lower()), tags=[str(i)]) for i, _d in enumerate(documents_prb)]
max_epochs = 20
vec_size = 20
alpha = 0.025
model = Doc2Vec(size=vec_size,
alpha=alpha,
min_alpha=0.00025,
min_count=1,
dm =1)
model.build_vocab(tagged_data)
for epoch in range(max_epochs):
print('iteration {0}'.format(epoch))
model.train(tagged_data,
total_examples=model.corpus_count,
epochs=model.iter)
# decrease the learning rate
model.alpha -= 0.0002
# fix the learning rate, no decay
model.min_alpha = model.alpha
model.save("d2v.model")
print("Model Saved")
model= Doc2Vec.load("d2v.model")
#to find the vector of a document which is not in training data
def doc2vec_score(s):
s_list = tokenize_and_stem(s)
v1 = model.infer_vector(s_list)
similar_doc = model.docvecs.most_similar([v1])
original_match = (X[int(similar_doc[0][0])])
score = similar_doc[0][1]
match = similar_doc[0][0]
return score,match
final_data = []
# df_ws is the list of 10K docs for which i want to find the similarity with above 20 docs
for index, row in df_ws.iterrows():
print(row['processed_description'])
data = (doc2vec_score(row['processed_description']))
L1=list(data)
L1.append(row['Number'])
final_data.append(L1)
with open('file_cosine_d2v.csv','w',newline='') as out:
csv_out=csv.writer(out)
csv_out.writerow(['score','match','INC_NUMBER'])
for row in final_data:
csv_out.writerow(row)
But, I am facing the strange issue, the results are highly un-reliable (Score is 0.9 even if there is not a slightest match) and score is changing with great margin every time. I am running the doc2vec_score function. Can someone please help me what is wrong here ?
First & foremost, try not using the anti-pattern of calling train multiple times in your own loop.
See this answer for more details: My Doc2Vec code, after many loops of training, isn't giving good results. What might be wrong?
If there's still a problem after that fix, edit your question to show the corrected code, and a more clear example of the output you consider unreliable.
For example, show the actual doc-IDs & scores, and explain why you think the probe document you're testing should be "not a slightest match" for any documents returned.
And note that if a document is truly nothing like the training documents, for example by using words that weren't in the training documents, it's not really possible for a Doc2Vec model to detect that. When it infers vectors for new documents, all unknown words are ignored. So you'll be left with a document using only known words, and it will return the best matches for that subset of your document's words.
More fundamentally, a Doc2Vec model is really only learning ways to contrast the documents that are in the universe demonstrated by the training set, by their words' cooccurrences. If presented with a document with either totally different words, or words whose frequencies/cooccurrences are totally unlike anything seen before, its output will be essentially random, without much meaningful relationship to other more-typical documents. (That'll be maybe-close, maybe-far, because in a way the training on the 'known universe' tends to fill the whole available space.)
So, you wouldn't want to use a Doc2Vec model trained only only positive examples of what you want to recognize, if you also want to recognize negative examples. Rather, include all kinds, then remember the subset that's relevant for certain in/out decisions – and use that subset for downstream comparisons, or multiple subsets to feed a more-formal classification or clustering algorithm.
Related
Currently I am using a csv file converted from a pcap and I took the column length from my csv file and used it as my embedding. The code compiles and I do get accuracy in the high 70s like 77 percent. I am just not sure if this is an appropriate choice for an embedding. I am also getting this issue of some data sets get weird results as UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in labels with no predicted samples. I know there is people who answered this question but I tried all of there methods and still no clue why my model works for some data sets and not all.Please if someone could confirm what I am doing makes sense or not that would really help me.
CSV file snapshot for reference
df['embeddings'] =df['Length']
embeddings = torch.from_numpy(df['embeddings'].to_numpy())
# normalizing degree values
scale = StandardScaler()
embeddings = scale.fit_transform(embeddings.reshape(-1,1))
I have two text datasets. Each dataset consists of multiple sequences and each sequence can have more than one sentence.
How do I measure if both datasets are from same distribution?
The purpose is to verify transfer learning from one distribution to another only if the difference between the distributions is statistically significant.
I am panning to use chi-square test but not sure if it will help for text data considering the high degrees of freedom.
update:
Example:
Supppose I want to train a sentiment classification model. I train a model on IMDb dataset and evaluate on IMDb and Yelp datasets. I found that my model trained on IMDb still does well on Yelp. But the question is how different these datasets are?
Train Dataset : https://www.kaggle.com/columbine/imdb-dataset-sentiment-analysis-in-csv-format?select=Train.csv
Eval 1: https://www.kaggle.com/columbine/imdb-dataset-sentiment-analysis-in-csv-format?select=Valid.csv
Eval 2: https://www.kaggle.com/omkarsabnis/sentiment-analysis-on-the-yelp-reviews-dataset
Now,
How different are train and eval 1?
How different are train and eval 2?
Is the dissimilarity between train and eval 2 by chance ? What is the statistical significance and p value?
The question "are text A and text B coming from the same distribution?" is somehow poorly defined. For example, these two questions (1,2) can be viewed as generated from the same distribution (distribution of all questions on StackExchange) or from different distributions (distribution of two different subdomains of StackExchange). So it's not clear what is the property that you want to test.
Anyway, you can come up with any test statistic of your choice, approximate its distribution in case of "single source" by simulation, and calculate the p-value of your test.
As a toy example, let's take two small corpora: two random articles from English Wikipedia. I'll do it in Python
import requests
from bs4 import BeautifulSoup
urls = [
'https://en.wikipedia.org/wiki/Nanjing_(Liao_dynasty)',
'https://en.wikipedia.org/wiki/United_States_Passport_Card'
]
texts = [BeautifulSoup(requests.get(u).text).find('div', {'class': 'mw-parser-output'}).text for u in urls]
Now I use a primitive tokenizer to count individual words in texts, and use root mean squared difference in word relative frequencies as my test statistic. You can use any other statistic, as long as you calculate it consistently.
import re
from collections import Counter
from copy import deepcopy
TOKEN = re.compile(r'([^\W\d]+|\d+|[^\w\s])')
counters = [Counter(re.findall(TOKEN, t)) for t in texts]
print([sum(c.values()) for c in counters])
# [5068, 4053]: texts are of approximately the same size
def word_freq_rmse(c1, c2):
result = 0
vocab = set(c1.keys()).union(set(c2.keys()))
n1, n2 = sum(c1.values()), sum(c2.values())
n = len(vocab)
for word in vocab:
result += (c1[word]/n1 - c2[word]/n2)**2 / n
return result**0.5
print(word_freq_rmse(*counters))
# rmse is 0.001178, but is this a small or large difference?
I get a value of 0.001178, but I don't know whether it's a large difference. So I need to simulate the distribution of this test statistic under the null hypothesis: when both texts are from the same distribution. To simulate it, I merge two texts into one, and then split them randomly, and calculate my statistic when comparing these two random parts.
import random
tokens = [tok for t in texts for tok in re.findall(TOKEN, t)]
split = sum(counters[0].values())
distribution = []
for i in range(1000):
random.shuffle(tokens)
c1 = Counter(tokens[:split])
c2 = Counter(tokens[split:])
distribution.append(word_freq_rmse(c1, c2))
Now I can see how unusual is the value of my observed test statistic under the null hypothesis:
observed = word_freq_rmse(*counters)
p_value = sum(x >= observed for x in distribution) / len(distribution)
print(p_value) # it is 0.0
print(observed, max(distribution), sum(distribution) / len(distribution)) # 0.0011 0.0006 0.0004
We see that when texts are from the same distribution, my test statistic is on average 0.0004 and almost never exceeds 0.0006, so the value of 0.0011 is very unusual, and the null hypothesis that two my texts originate from the same distribution should be rejected.
I wrote an article which is similar to your problem but not exactly the same.
https://towardsdatascience.com/a-new-way-to-bow-analysis-feature-engineering-part1-e012eba90ef
The problem that I was trying to solve is to check if a word has different (significant) distributions across categories or labels.
There are a few similarities between your problem and the one I had mentioned above.
You want to compare two sources of datasets, which can be taken as two different categories
Also, to compare the data sources, you will have to compare the words as sentences can't be directly compared
So, my proposed solution to this will be as:
Create words features across the two datasets using count-vectorizer and get top X words from each
Let's say you have total distinct words as N, now initialize count=0 and start to compare the distribution for each word and if the differences are significant increment the counter. Also, there could be cases where a word only exists in one of the datasets and that is a good new, by that I mean it shows that it is a distinguishing feature, so, for this also increment the count
Let's say the total count is n. Now, the lower is the n/N ratio, similar two texts are and vice-a-versa
Also, to verify this methodology - Split the data from a single source into two (random sampling) and run the above analysis, if the n/N ratio is closer to 0 which indicates that the two data sources are similar which also is the case.
Please let me know if this approach worked or not, also if you think there are any flaws in this, I would love to think and try evolving it.
I am a newbie in machine Learning, i am building a complaint categorizer and i want to provide a feedback model so that it can improve over time
import numpy
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.naive_bayes import MultinomialNB
value=[
'drought',
'robber',
]
targets=[
'water_department',
'police_department',
]
classifier = MultinomialNB()
vectorizer = CountVectorizer()
counts = vectorizer.fit_transform(value)
classifier.partial_fit(counts[:1], targets[:1],classes=numpy.unique(targets))
for c,t in zip(counts[1:],targets[1:]):
classifier.partial_fit(c, t.split())
value.append('dogs') #new value to train
targets.append('animal_department') #new target
vectorize = CountVectorizer()
counts = vectorize.fit_transform(value)
print counts
print targets
print vectorize.vocabulary_
####problem lies here
classifier.partial_fit(counts["""dont know the index of new value"""], targets[-1])
####problem lies here
Even if i somehow find the index of newly inserted value, it is giving the error
ValueError: Number of features 3 does not match previous data 2.
even thought i made it to insert one value at a time
I will try to answer the question from a general point of view. There are two sources of problem in the Naive Bayes (NB) approach described here:
Out-of-vocabulary (OOV) problem
Incremental training of NB
OOV problem: The simplest way to tackle the OOV problem is to decompose every word into character 3 grams. How many such 3-grams are possible? Assuming lower-casing there are only 26 possible ways to fill each place and hence the total number of possible character 3-grams is 26^3=17576, which is significantly lower than the number of possible English words that you're likely to see in text.
Hence, generally speaking, while training NB, a good idea is to use probabilities of character n-grams (n=3,4,5). This will drastically reduce the OOV problem.
Incremental training: For incremental training, given a new sentence decompose it into terms (character n-grams). Update the count of of each term for its corresponding observed class label. For example, if count(t,c) denotes how many times was the term t observed in class c, simply update the count if you see t in class 0 (or class 1) during incremental training. Updating the counts will update the maximum likelihood probability estimates as well.
I'm setting up the new Tensorflow Object Detection API to find small objects in large areas of satellite imagery. It works quite well - it finds all 10 objects I want, but I also get 50-100 false positives [things that look a little like the target object, but aren't].
I'm using the sample config from the 'pets' tutorial, to fine-tune the faster_rcnn_resnet101_coco model they offer. I've started small, with only 100 training examples of my objects (just 1 class). 50 examples in my validation set. Each example is a 200x200 pixel image with a labeled object (~40x40) in the center. I train until my precision & loss curves plateau.
I'm relatively new to using deep learning for object detection. What is the best strategy to increase my precision? e.g. Hard-negative mining? Increase my training dataset size? I've yet to try the most accurate model they offer faster_rcnn_inception_resnet_v2_atrous_coco as i'd like to maintain some speed, but will do so if needed.
Hard-negative mining seems to be a logical step. If you agree, how do I implement it w.r.t setting up the tfrecord file for my training dataset? Let's say I make 200x200 images for each of the 50-100 false positives:
Do I create 'annotation' xml files for each, with no 'object' element?
...or do I label these hard negatives as a second class?
If I then have 100 negatives to 100 positives in my training set - is that a healthy ratio? How many negatives can I include?
I've revisited this topic recently in my work and thought I'd update with my current learnings for any who visit in the future.
The topic appeared on Tensorflow's Models repo issue tracker. SSD allows you to set the ratio of how many negative:postive examples to mine (max_negatives_per_positive: 3), but you can also set a minimum number for images with no postives (min_negatives_per_image: 3). Both of these are defined in the model-ssd-loss config section.
That said, I don't see the same option in Faster-RCNN's model configuration. It's mentioned in the issue that models/research/object_detection/core/balanced_positive_negative_sampler.py contains the code used for Faster-RCNN.
One other option discussed in the issue is creating a second class specifically for lookalikes. During training, the model will attempt to learn class differences which should help serve your purpose.
Lastly, I came across this article on Filter Amplifier Networks (FAN) that may be informative for your work on aerial imagery.
===================================================================
The following paper describes hard negative mining for the same purpose you describe:
Training Region-based Object Detectors with Online Hard Example Mining
In section 3.1 they describe using a foreground and background class:
Background RoIs. A region is labeled background (bg) if its maximum
IoU with ground truth is in the interval [bg lo, 0.5). A lower
threshold of bg lo = 0.1 is used by both FRCN and SPPnet, and is
hypothesized in [14] to crudely approximate hard negative mining; the
assumption is that regions with some overlap with the ground truth are
more likely to be the confusing or hard ones. We show in Section 5.4
that although this heuristic helps convergence and detection accuracy,
it is suboptimal because it ignores some infrequent, but important,
difficult background regions. Our method removes the bg lo threshold.
In fact this paper is referenced and its ideas are used in Tensorflow's object detection losses.py code for hard mining:
class HardExampleMiner(object):
"""Hard example mining for regions in a list of images.
Implements hard example mining to select a subset of regions to be
back-propagated. For each image, selects the regions with highest losses,
subject to the condition that a newly selected region cannot have
an IOU > iou_threshold with any of the previously selected regions.
This can be achieved by re-using a greedy non-maximum suppression algorithm.
A constraint on the number of negatives mined per positive region can also be
enforced.
Reference papers: "Training Region-based Object Detectors with Online
Hard Example Mining" (CVPR 2016) by Srivastava et al., and
"SSD: Single Shot MultiBox Detector" (ECCV 2016) by Liu et al.
"""
Based on your model config file, the HardMinerObject is returned by losses_builder.py in this bit of code:
def build_hard_example_miner(config,
classification_weight,
localization_weight):
"""Builds hard example miner based on the config.
Args:
config: A losses_pb2.HardExampleMiner object.
classification_weight: Classification loss weight.
localization_weight: Localization loss weight.
Returns:
Hard example miner.
"""
loss_type = None
if config.loss_type == losses_pb2.HardExampleMiner.BOTH:
loss_type = 'both'
if config.loss_type == losses_pb2.HardExampleMiner.CLASSIFICATION:
loss_type = 'cls'
if config.loss_type == losses_pb2.HardExampleMiner.LOCALIZATION:
loss_type = 'loc'
max_negatives_per_positive = None
num_hard_examples = None
if config.max_negatives_per_positive > 0:
max_negatives_per_positive = config.max_negatives_per_positive
if config.num_hard_examples > 0:
num_hard_examples = config.num_hard_examples
hard_example_miner = losses.HardExampleMiner(
num_hard_examples=num_hard_examples,
iou_threshold=config.iou_threshold,
loss_type=loss_type,
cls_loss_weight=classification_weight,
loc_loss_weight=localization_weight,
max_negatives_per_positive=max_negatives_per_positive,
min_negatives_per_image=config.min_negatives_per_image)
return hard_example_miner
which is returned by model_builder.py and called by train.py. So basically, it seems to me that simply generating your true positive labels (with a tool like LabelImg or RectLabel) should be enough for the train algorithm to find hard negatives within the same images. The related question gives an excellent walkthrough.
In the event you want to feed in data that has no true positives (i.e. nothing should be classified in the image), just add the negative image to your tfrecord with no bounding boxes.
I think I was passing through the same or close scenario and it's worth it to share with you.
I managed to solve it by passing images without annotations to the trainer.
On my scenario I'm building a project to detect assembly failures from my client's products, at real time.
I successfully achieved very robust results (for production env) by using detection+classification for components that has explicity a negative pattern (e.g. a screw that has screw on/off(just the hole)) and only detection for things that doesn't has the negative pattens (e.g. a tape that can be placed anywhere).
On the system it's mandatory that the user record 2 videos, one containing the positive scenario and another containing the negative (or the n videos, containing n patterns of positive and negative so the algorithm can generalize).
After a while testing I found out that if I register to detected only tape the detector was giving very confident (0.999) false positive detections of tape. It was learning the pattern where the tape was inserted instead of the tape itself. When I had another component (like a screw on it's negative format) I was passing the negative pattern of tape without being explicitly aware of it, so the FPs didn't happen.
So I found out that, in this scenario, I had to necessarily pass the images without tape so it could differentiate between tape and no-tape.
I considered two alternatives to experiment and try to solve this behavior:
Train passing an considerable amount of images that doesn't has any annotation (10% of all my negative samples) along with all images that I have real annotations.
On the images that I don't have annotation I create a dummy annotation with a dummy label so I could force the detector to train with that image (thus learning the no-tape pattern). Later on, when get the dummy predictions, just ignore them.
Concluded that both alternatives worked perfectly on my scenario.
The training loss got a little messy but the predictions work with robustness for my very controlled scenario (the system's camera has its own box and illumination to decrease variables).
I had to make two little modifications for the first alternative to work:
All images that didn't had any annotation I passed a dummy annotation (class=None, xmin/ymin/xmax/ymax=-1)
When generating the tfrecord files I use this information (xmin == -1, in this case) to add an empty list for the sample:
def create_tf_example(group, path, label_map):
with tf.gfile.GFile(os.path.join(path, '{}'.format(group.filename)), 'rb') as fid:
encoded_jpg = fid.read()
encoded_jpg_io = io.BytesIO(encoded_jpg)
image = Image.open(encoded_jpg_io)
width, height = image.size
filename = group.filename.encode('utf8')
image_format = b'jpg'
xmins = []
xmaxs = []
ymins = []
ymaxs = []
classes_text = []
classes = []
for index, row in group.object.iterrows():
if not pd.isnull(row.xmin):
if not row.xmin == -1:
xmins.append(row['xmin'] / width)
xmaxs.append(row['xmax'] / width)
ymins.append(row['ymin'] / height)
ymaxs.append(row['ymax'] / height)
classes_text.append(row['class'].encode('utf8'))
classes.append(label_map[row['class']])
tf_example = tf.train.Example(features=tf.train.Features(feature={
'image/height': dataset_util.int64_feature(height),
'image/width': dataset_util.int64_feature(width),
'image/filename': dataset_util.bytes_feature(filename),
'image/source_id': dataset_util.bytes_feature(filename),
'image/encoded': dataset_util.bytes_feature(encoded_jpg),
'image/format': dataset_util.bytes_feature(image_format),
'image/object/bbox/xmin': dataset_util.float_list_feature(xmins),
'image/object/bbox/xmax': dataset_util.float_list_feature(xmaxs),
'image/object/bbox/ymin': dataset_util.float_list_feature(ymins),
'image/object/bbox/ymax': dataset_util.float_list_feature(ymaxs),
'image/object/class/text': dataset_util.bytes_list_feature(classes_text),
'image/object/class/label': dataset_util.int64_list_feature(classes),
}))
return tf_example
Part of the traning progress:
Currently I'm using tensorflow object detection along with tensorflow==1.15, using faster_rcnn_resnet101_coco.config.
Hope it will solve someone's problem as I didn't found any solution on the internet. I read a lot of people telling that faster_rcnn is not adapted for negative training for FPs reduction but my tests proved the opposite.
This my sound as very naive question. I checked on google and many YouTube videos for beginners and pretty much, all explain data weighting as something the most obvious. I still do not understand why data is being weighted.
Let's assume I have four features:
a b c d
1 2 1 4
If I pass each value to Sigmond function, I'll receive -1 >< 1 value already.
I really don't understand why data needs or it is recommended to be weighted first. If you could explain to me this in very simple manner, I would appreciate it a lot.
I think you are not talking about weighing data but features.
A feature is a column in your table and as data I would understand rows.
The confusion comes now from the fact that weighing rows is also sometimes sensible, e.g., if you want to punish misclassification of positive class more.
Why do we need to weigh features?
I assume you are talking about a modle like
prediction = sigmoid(sum_i weight_i * feature_i) > base
Let's assume you want to predict whether a person is overweight based on Bodyweight, height, and age.
In R we can generate a sample dataset as
height = rnorm(100,1.80,0.1) #normal distributed mean 1.8,variance 0.1
weight = rnorm(100,70,10)
age = runif(100,0,100)
ow = weight / (height**2)>25 #overweight if BMI > 25
data = data.frame(height,weight,age,bc,ow)
if we now plot the data you can see that at least my sample of the data can be separated with a straight line in weight/height. However, age does not provide any value. If we weight it prior to the sum/sigmoid you can put all factors into relation.
Furthermore, as you can see from the following plot the weight/height have a very different domain. Hence, they need to be put into relation, such that the line in the following plot has the right slope, as the value of weight have are one order of magnitude larger